evapotranspiration estimation: a study of methods in the
TRANSCRIPT
Utah State University Utah State University
DigitalCommons@USU DigitalCommons@USU
All Graduate Theses and Dissertations Graduate Studies
5-2016
Evapotranspiration Estimation: A Study of Methods in the Western Evapotranspiration Estimation: A Study of Methods in the Western
United States United States
Clayton S. Lewis Utah State University
Follow this and additional works at: https://digitalcommons.usu.edu/etd
Part of the Civil Engineering Commons
Recommended Citation Recommended Citation Lewis, Clayton S., "Evapotranspiration Estimation: A Study of Methods in the Western United States" (2016). All Graduate Theses and Dissertations. 4683. https://digitalcommons.usu.edu/etd/4683
This Dissertation is brought to you for free and open access by the Graduate Studies at DigitalCommons@USU. It has been accepted for inclusion in All Graduate Theses and Dissertations by an authorized administrator of DigitalCommons@USU. For more information, please contact [email protected].
.
EVAPOTRANSPIRATION ESTIMATION: A STUDY OF METHODS
IN THE WESTERN UNITED STATES
by
Clayton S. Lewis
A dissertation submitted in partial fulfillment of the requirements of the degree
of
DOCTOR OF PHILOSOPHY
in
Civil Engineering
Approved: _________________________ Christopher M. U. Neale Major Professor _________________________ Robert W. Hill Committee Member _________________________ Jagath J. Kaluarachchi Committee Member
_________________________ Gilberto E. Urroz Committee Member _________________________ Joanna Endter-Wada Committee Member _________________________ Mark McLellan Vice President for Research and Dean of the School of Graduate Studies
UTAH STATE UNIVERSITY
Logan, Utah
2016
iii
ABSTRACT
Evapotranspiration Estimation: A Study of Methods
in the Western United States
by
Clayton S. Lewis, Doctor of Philosophy
Utah State University, 2016 Major Professor: Dr. Christopher M. U. Neale Department: Civil and Environmental Engineering
This research focused on estimating evapotranspiration (i.e., the amount
of water vaporizing into the atmosphere through processes of surface
evaporation and plant transpiration) under both theoretical and actual conditions.
There were two study areas involved: one, on a large scale where 704
agriculturally-representative, electronic weather stations were used to evaluate
the drivers and calculated reference evapotranspiration of a NASA gridded
weather forcing model in the 17 western states in the contiguous U.S.; and two,
transpiration of invasive saltcedar (Tamarix sp.) in the floodplain of the lower
Colorado River, California, with Bowen ratio, eddy covariance, and groundwater
fluxes. In this study, a fire destroyed the saltcedar forest, which allowed
comparison of evapotranspiration before and after\ this event.
iv
Comparison of the input weather parameters showed some variance
between the electronic weather stations and the gridded model, but calculated
reference evapotranspiration performed well by relying on the better input and
more highly weighed variables of air temperature and downward shortwave
radiation. Only in the southern portions of California, Arizona, and New Mexico
were the evapotranspiration estimates using the gridded dataset not well
correlated with the electronic weather stations and not recommended for
prediction. Saltcedar evapotranspiration was found to match more recent and
conservative estimates for the phreatophyte than what was historically portrayed
in the literature. Horizontal advection from the surrounding desert was also
observed to affect the riparian energy balance. Annual average total
evapotranspiration before a fire varied from 0.60-1.44 meters/year to 0.25-1.00
meters/year post fire.
(132 pages)
v
PUBLIC ABSTRACT
Evapotranspiration Estimation: A Study of Methods
in the Western United States
Clayton S. Lewis Theoretical water use of well-watered vegetation in the western United
States was investigated by comparing a gridded dataset developed from satellite
and ground instruments to weather stations representative of irrigated agricultural
conditions. Since wetter environments are cooler and therefore subject to lower
levels of evaporation than the same scenario with warmer temperatures, models
derived from dryland and often populated areas overestimate potential plant
needs in semi-arid or arid conditions. Evaluation of the model revealed an
acceptable fit for air temperatures and solar radiation but with less confidence in
humidity and wind speeds. Ultimately, the last two parameters were minor
components in the calculation of reference evapotranspiration, which performed
acceptably well at an hourly time step except for some high biases in states
along the southern U.S. border. These findings substantiate application of the
model for calculation of plant water requirements 1979-present in most areas.
Actual water use of invasive saltcedar was studied along the Colorado
River in a wildlife refuge south of Palo Verde, California. Surface meteorological
towers and groundwater wells measured water fluxes from 2006 until a fire
burned the refuge and most of the equipment in August 2011. New instruments
were installed and measured the evapotranspiration of the regrowth. Change in
evapotranspiration was evidenced and models verified by a zeroed state. This
control state of no (or nearly no) vegetation was very beneficial in this study.
Saltcedar water use was assessed against other studies and the theoretical
estimates mentioned above and found to be comparable to levels required by
native vegetation.
vi
ACKNOWLEDGMENTS
I take no credit for the gain in knowledge I have received while
researching or for what is recorded in this dissertation but attribute every
enlightened thought followed by incredible fruition to my Father in Heaven.
There have been many times while sitting at my desk engaged in programming,
writing, or reading an article when unmistakable inspiration has whispered a
thought into my mind. Frequently, this occurred in the final moments of a day or
nearing the end of a project deadline after I had put in a full effort. Grateful I am
for the God who has miraculously sustained me throughout my studies at Utah
State University. I do accept responsibility for any error contained herein.
To my wife’s patience and unbeleaguered support during my years of
graduate studies and bearing of three children: Thank you, and I love you.
To Dr. Hill, my first advisor, thanks for just scratching your head when I
shared some big ideas and for instilling an engineering mindset.
To Dr. Neale, thanks for allowing me to soar untethered into the heights of
success.
To Drs. Endter-Wada and Kjelgren, thanks for the diversion coding
WaterMAPS and for the experience working with water agencies.
To Burdette Barker, Ashish Masih, Ian Ferguson, Diana Glenn, Hatim
Geli, and Marcelo Biudes, thanks for your association and help.
Clayton Lewis
vii
CONTENTS
ABSTRACT .......................................................................................................... iii
PUBLIC ABSTRACT ............................................................................................ v
ACKNOWLEDGMENTS .......................................................................................vi
CONTENTS ......................................................................................................... vii
LIST OF TABLES ................................................................................................. x
LIST OF FIGURES ...............................................................................................xi
CHAPTER
1. INTRODUCTION ....................................................................................... 1
1.1 Overview ................................................................................... 1
1.1.1 Purpose ....................................................................... 1 1.1.2 Reference Evapotranspiration ..................................... 3 1.1.3 Saltcedar Evapotranspiration ...................................... 5 1.1.4 Research Objectives ................................................... 7
1.2 Methodology .............................................................................. 8
1.2.1 Penman-Monteith Reference Evapotranspiration ...... 10 1.2.2 Bowen Ratio Energy Balance Technique .................. 12 1.2.3 Eddy Covariance Technique ..................................... 14 1.2.4 Groundwater Fluxes .................................................. 15 1.2.5 Remote Sensing Energy Balance Approach ............. 17
1.3 References .............................................................................. 18
2. COMPARISON OF THE NLDAS WEATHER FORCING MODEL TO AGROMETEOROLOGICAL MEASUREMENTS IN THE WESTERN UNITED STATES .................................................................. 23
2.1 Abstract ................................................................................... 23 2.2 Introduction ............................................................................. 24 2.3 Background ............................................................................. 24 2.4 Data Preparation ..................................................................... 28
viii
2.4.1 Retrieval .................................................................... 28 2.4.2 Weather Station Locale Filter .................................... 30
2.5 Comparison of NLDAS Data ................................................... 35
2.5.1 Weather Parameters ................................................. 35 2.5.2 Reference ET ............................................................ 38
2.6 Conclusions............................................................................. 43 2.7 Acknowledgments ................................................................... 43 2.8 References .............................................................................. 44
3. BOWEN RATIO AND EDDY COVARIANCE SYSTEM PERFORMANCE OVER SALTCEDAR UNDER DIFFERENT ATMOSPHERIC STABILITY AND ADVECTION CONDITIONS .............. 48
3.1 Abstract ................................................................................... 48 3.2 Introduction ............................................................................. 48 3.3 Theoretical Foundations .......................................................... 50 3.4 Methodology Limitations ......................................................... 52 3.5 Advection and Atmospheric Stability ....................................... 56 3.6 Site Characteristics ................................................................. 57 3.7 Data Analysis .......................................................................... 60 3.8 Results and Discussion ........................................................... 62 3.9 Summary and Conclusions ..................................................... 68 3.10 Acknowledgments ................................................................. 68 3.11 References ............................................................................ 69
4. SALTCEDAR EVAPOTRANSPIRATION PRE AND POST FIRE ............ 73
4.1 Abstract ................................................................................... 73 4.2 Introduction ............................................................................. 73 4.3 Methods .................................................................................. 75 4.4 Materials ................................................................................. 77 4.5 Data Analysis and Results ...................................................... 82 4.6 Discussion ............................................................................... 92 4.7 Conclusions............................................................................. 95 4.8 Acknowledgments ................................................................... 96 4.9 References .............................................................................. 96
5. SUMMARY AND CONCLUSIONS ......................................................... 101
5.1 Objective Fulfillment .............................................................. 102 5.2 Conclusions........................................................................... 105 5.3 References ............................................................................ 105
ix
APPENDICES .................................................................................................. 106
Appendix A: Permissions ............................................................ 107 Appendix B: Curriculum Vitae ..................................................... 116
x
LIST OF TABLES
Table Page
2.1. Electronic Weather Station Networks over Agricultural Areas for 17 States in the Western United States ........................................................ 27
2.2. Sequential Steps Taken to Programmatically Filter and Correct Electronic Weather Station Data .............................................................. 34
3.1. Advection-Corrected to Non-Corrected Fraction of Seasonal Sensible and Latent Heat Fluxes over Salt Cedar in Cibola, California, Pre and Post Fire .................................................................... 67
4.1. Averaged Groundwater Properties for Multiple Wells at 3 Sites in the Cibola National Wildlife Refuge Pre and Post Fire ............................. 91
4.2. Average Annual Groundwater ET from Multiple Wells at 3 Sites in the Cibola National Wildlife Refuge Pre and Post Fire Assuming Conservative Estimates of Specific Yield (meter/year) ............................ 92
xi
LIST OF FIGURES
Figure Page
2.1. Annual Count of Operational Agricultural Electronic Weather Stations in the Western United States (1981-2012) ................................. 27
2.2. 704 Agricultural Electronic Weather Stations Superimposed over the NASS 2012 Cropland Data Layer in the Western United States ........ 32
2.3. NLDAS and Agricultural Weather Station Scatter Plots of Hourly Solar Radiation, Air Temperature, Wind Speed, and Relative Humidity at Four Irrigated Locations in the Western United States .......... 36
2.4. IDW Interpolated R Squared of Hourly Solar Radiation, Air Temperature, Wind Speed, and Relative Humidity of the Comparison between 704 Agricultural Weather Stations and NLDAS in the Western United States ...................................................... 39
2.5. IDW Interpolated Bias of Hourly Solar Radiation, Air Temperature, Wind Speed, and Relative Humidity of the Comparison between 704 Agricultural Weather Stations and NLDAS in the Western United States ........................................................................................... 40
2.6. IDW Interpolated Root Mean Squared Error of Hourly Solar Radiation, Air Temperature, Wind Speed, and Relative Humidity of the Comparison between 704 Agricultural Weather Stations and NLDAS in the Western United States ...................................................... 41
2.7. IDW Interpolated R Squared, Bias, and Root Mean Squared Error of Hourly Short or Grass Reference Evapotranspiration of the Comparison between 704 Agricultural Weather Stations and NLDAS in the Western United States ...................................................... 42
3.1. Location of Three Surface Flux Measurement Sites over a Salt Cedar Forest within the Cibola Wildlife National Refuge near the Colorado River ......................................................................................... 59
3.2. Measurement Towers and Surrounding Salt Cedar Forest Pre and Post Fire at the Cibola National Wildlife Refuge [(A) Diablo Tower Pre Fire, 2008 (B) Slithern Tower Pre Fire, 2008 (C) SlithernTower After Fire, 2012 (D) Rebuilt Slithern Tower Post Fire, 2013] .................... 60
xii
3.3. Hourly Fraction of Bowen Ratios obtained from Eddy Covariance and Bowen Ratio Energy Balance Instrumentation over Salt Cedar at Diablo Station in Cibola, California, Pre Fire (2008-2009) ................... 63
3.4. Hourly Fraction of Bowen Ratios obtained from Eddy Covariance and Bowen Ratio Energy Balance Instrumentation over Salt Cedar at Slithern Station in Cibola, California, Post Fire (2012-2013) ................ 64
3.5. Average Hourly Fraction of Bowen Ratios obtained from Eddy Covariance and Bowen Ratio Energy Balance Instrumentation over Salt Cedar in Cibola, California, Pre and Post Fire (2008-2013) .............. 65
3.6. Average Hourly Surface Fluxes obtained from Bowen Ratio Energy Balance Instrumentation over Salt Cedar at Diablo Station in Cibola, California, Pre Fire (2008-2009) .................................................. 66
3.7. Average Hourly Surface Fluxes obtained from Bowen Ratio Energy Balance Instrumentation over Salt Cedar at Slithern Station in Cibola, California, Post Fire (2012-2013) ................................................. 66
4.1. False Color (Infrared, Red, Green) Images from the National Agriculture Imagery Program (NAIP) of the Tower Sites in the Cibola National Wildlife Refuge, (A) June 26, 2005; (B) May 25, 2009; (C) April 26 and 28, 2010; (D) April 23, 2012; (E) May 16, 2014 ......................................................................................................... 79
4.2. Tower and Groundwater Well Observation Locations within the Cibola National Wildlife Refuge ............................................................... 81
4.3. Saltcedar Regrowth in the Cibola National Wildlife Refuge Post Fire (From Top to Bottom, March 12, 2012; May 29, 2013; August 20, 2013)........................................................................................................ 82
4.4. Weekly Averaged Daily ET from Bowen Ratio Systems in the Cibola National Wildlife Refuge, (Top: 2006-2010; Bottom: 2012-2013)........................................................................................................ 84
4.5. Weekly Averaged Daily ET from Eddy Covariance Systems in the Cibola National Wildlife Refuge, (Top: 2007-2009; Bottom: 2012-2013)........................................................................................................ 86
4.6. Weekly Average Daily Groundwater Surface Elevation from Multiple Wells at 3 Sites in the Cibola National Wildlife Refuge ............................ 88
xiii
4.7. Example of 24 Hour Fourier-White Method Implementation for 3 Days at Diablo Tower Well, Cibola National Wildlife Refuge (July 1-7, 2011) .................................................................................................... 88
4.8. Weekly Averaged Daily Groundwater ET from Multiple Wells at 3 Sites in the Cibola National Wildlife Refuge, (From Top to Bottom, 2006-2010; 2011; 2012-2013) ................................................................. 91
1
CHAPTER 1
INTRODUCTION
1.1 Overview
1.1.1 Purpose
Access to, acquisition of, and allocation of potable and non-potable water
has been the motivation for colossal civil projects, numerous governmental pacts
and entities, and disputes at local and state levels. Where deficient supplies of
fresh water exist, oversight becomes even more intense. This accurately
describes conditions prevalent in the Great American Desert or western United
States. Water storage and delivery systems being essential in this arid
environment, citizens and communities have been obliged to operate jointly this
natural resource. Accordingly, development of better management practices has
been sought after to ensure equity of use. With academia thus induced, higher
institutions have answered the call with Utah State University in Logan, Utah,
being a forerunner in the field of water management and engineering. This work
sought to complement the backdrop of accomplishments by increasing
knowledge of evapotranspiration in the West. Contributors and partners in this
study were Utah State University, U.S. Bureau of Reclamation, U.S. Geological
Survey, U.S. Fish and Wildlife Service, National Oceanic and Atmospheric
Administration, National Air and Space Administration, Palo Verde Irrigation
District, Six Agencies of the Lower Colorado River, various universities across
the United States, and many organizations collecting weather data.
2
For decades, the consumptive use of water by vegetation in agricultural
and riparian lands has been studied and estimates published. As the demand
per unit of water continues to increase, managing agencies must account for
every transaction accurately or at least much closer than past rough estimates.
Currently, a multitude of researchers throughout the world are actively striving to
resolve these issues by developing and installing more advanced instrumentation
to increase temporal and spatial resolution of field measurements and by
developing new methods to assimilate datasets to create comprehensive models
of the hydrological cycle. In an effort to broaden documented knowledge of
processes pertaining to vegetative transpiration and surface evaporation, this
study investigated these three questions:
(1) Can regional reference evapotranspiration be modeled by a gridded
national weather forcing dataset in the 17-state American West while
satisfactorily agreeing with an extensive set of ground-based weather
stations?
(2) Does the Bowen ratio method of evapotranspiration estimation
produce accurate results in an advective, desert environment?
(3) What are the effects of a fire on a riparian stretch of saltcedar and the
change in evapotranspiration pre and post fire on the groundwater fluxes?
3
1.1.2 Reference Evapotranspiration
With the possibility of an enormous variety of crops and vegetation being
grown even in a specific microclimate (which are also subject to fertilizers,
irrigation, diseases, drought, variable topography and soil, and differences in
species), the task of determining consumptive use on these vegetated land
surfaces is a daunting one. Available water and nutrients, crop phenology, and
evaporative potential all need to be tracked over non-dormant periods. Since
actual evapotranspiration is currently impossible to directly measure for the
unique location and characteristics of each land cover, empirical methods to
estimate evapotranspiration—based on theory to some degree—have been
proposed. The common approach has been to account for site irregularities or
crop particularities in reductions to or additions to a single reference crop.
According to Allen et al. (1998), reference crop evapotranspiration is defined by a
“standardized vegetative surface” that reflects the “evaporation power of the
atmosphere” of observed climatic conditions. Crop growth stages, including
season initiation and termination, are described by crop coefficients (Jensen,
1968) that are the ratio of a given crop evapotranspiration to that of the estimated
reference evapotranspiration. These crop coefficients may indicate either actual
evapotranspiration or potential evapotranspiration, the latter being “the amount of
water that would be evaporated and transpired if it were available” (Thornthwaite,
1947) to which further penalties could be applied indicating actual substandard
conditions.
4
Many researchers have recommended methods to calculate reference
evapotranspiration from surface measurements which can be grouped into one of
four categories based on theory incorporated: energy balance, mass transfer, an
empirical relationship to an observed measurement, or a combination of these.
Equations such as Penman (1948), Thornthwaite (1948), Makkink (1957), Turc
(1961), Blaney and Criddle (1962), Jensen and Haise (1963), Monteith (1965),
Christiansen (1968), Priestley and Taylor (1972), Pruitt and Doorenbos (1977),
and Hargreaves et al. (1985) are examples of methods that have been applied
extensively in this discipline and results compared to disturbed point
measurements and to each other. Selection of an evapotranspiration equation
depends on weather parameters in existing datasets, their frequency, and the
overall site climate. Some equations have proven more accurate than others
with modified versions of the combination Penman-Monteith equation performing
the best worldwide (Allen et al. 1998).
Calculation of the water requirement for a given crop coincides with a
reference evapotranspiration equation, which, historically, has been the Blaney-
Criddle equation due to its reduced input of monthly average temperature and
site calibrations (Allen et al. 1998). However, enhanced spatial and temporal
estimates of reference evapotranspiration are being developed with the advent of
new technologies: the electronic weather station; increased data storage,
sampling, transfer rates, and processing capacity; better sensors; satellite
imagery; and improved modeling techniques. At present, separate standards
based on divergent and now obsolete methodologies are administered by each
5
state, which inherently poses an obstacle in shared basins. A single method to
calculate reference evapotranspiration for the whole region with tolerable
accuracy would alleviate this problem and function as a more equitable allocation
metric.
1.1.3 Saltcedar Evapotranspiration
Saltcedar, also known by the genus Tamarix, is a dense shrub or tree
native to Europe, Asia, and Africa (Baum 1978), and several species from
Eurasia were imported into the eastern United States in the nineteenth century to
build wind breaks, control erosion along stream channels, produce shade, and be
planted as ornamentals. Like the name suggests, saltcedar tolerates saline
environments and also retains salt in the glands of its tiny leaves. Upon leaf fall,
the surface is temporally salted, giving an advantage in outcompeting rival
vegetative growth. Saltcedar also exhibits further survival traits including
resistance to flooding and droughts, susceptibility to wildfires contrasted with
quick regrowth from the root crown after a burn, and propagation through high
seed production and vegetative reproduction. Once introduced into the arid and
semi-arid American West from Eastern nurseries, saltcedar proliferated along
river corridors and supplanted local riparian vegetation such as cottonwoods,
willows, and mesquite. This invasive plant is now targeted with successive
mechanical and biological mechanisms for removal. (Di Tomaso 1998, Glenn
and Nagler 2005)
6
In addition, saltcedar has been condemned for excessive consumption of
groundwater, a safeguarded commodity in the western states. Stromberg et al.
(2009) at length summarized perceptions of saltcedar water requirements by the
public, scientists, and officials over the past 60 years. Political pressure to
eradicate what was viewed as an invasive and useless pest may have biased
initial, often indirectly-derived approximations of water use, which overestimated
2-3 times what has been measured the past decade with improved methods.
These high estimates have been touted as the reason for saltcedar removal.
However, studies have not shown anticipated water savings with saltcedar
elimination, which could be due to a number of groundwater recharge factors
(Shafroth et al. 2005). Nevertheless, saltcedar at present is considered an
invasive weed using significant quantities of water. It, perhaps, matches more
closely native riparian vegetation in water consumption but outcompetes and
grows in more extensive forests in the floodplain of western U.S. river systems.
Many factors regulate saltcedar water consumption, which may include elevation,
latitude, climate, groundwater levels, drought, salinity, growth stages, and
competition. The effects of these factors on saltcedar evapotranspiration are not
fully understood, and wide-scale estimates are still reliable only within a probable
range.
Evapotranspiration of saltcedar in the southwestern U.S. has been
measured by Bowen ratio towers, eddy covariance systems, scintillometers, and
estimated using remote sensing, information from groundwater wells, lysimetry,
and sap flow techniques (Cleverly et al. 2002, Nagler et al. 2009, Taghvaeian
7
2011, Geli et al. 2012). Each of these methods contains intrinsic errors that can
be enlarged through user mismanagement and together yield uncertainties to
well over 100 percent (Sher and Quigley 2013). In the case of the Bowen ratio
energy balance, import of energy by advection can invalidate results (Perez et al.
1999) but can be verified against the more robust eddy covariance method,
although itself limited by stable atmospheric conditions and wind direction.
Tracking groundwater fluctuations will accurately illustrate withdrawals from the
saturated layer but not the magnitude, whereas remote sensing approaches can
suffer from lower temporal resolution and proper calibration. Thus, refining
accurate estimates of saltcedar water use has proved challenging and requires
careful collection and evaluation of data (Westenburg et al. 2006).
1.1.4 Research Objectives
Further research in the fields of reference evapotranspiration and riparian
saltcedar water requirements would lead to new and pertinent understandings
affecting policies both political and academic. In an attempt to answer the
previous inquiries, the following research aims were evaluated:
(1) Comparison of the North American Land Data Assimilation System
(NLDAS-2, 2013) to agricultural electronic weather station networks in the
American West for the period 1981-2012 on an hourly time step for the
meteorological parameters of solar radiation, air temperature, wind speed,
and relative humidity as well as calculated reference evapotranspiration.
8
(2) Establishment of actual evapotranspiration of saltcedar in the
advective, desert climate of the Cibola National Wildlife Refuge (CNWR,
2013) along the Lower Colorado River by analyzing groundwater, Bowen
ratio, eddy covariance, and remote sensing datasets before and after a
decimating wildfire in 2011.
Both objectives required original research and expanded scientific
knowledge in the fields of reference and saltcedar evapotranspiration. Objective
1 referred to the initial question of comparing a national gridded dataset to
ground observations, while Objective 2 satisfied the second and third questions
in Section 1.1.1 by affording multiple energy flux datasets in an advective and
arid environment overrun by saltcedar that was recently burned. Following a
format of standalone journal papers, these objectives and the underlying
questions are specifically studied in Chapters 2-4.
1.2 Methodology
In an age with science and technology advancing at a rapid rate, it may be
no surprise in the near future to be able to track or correctly determine all of
Earth’s hydrologic movements. However, current models suffer from incongruent
subsampling in representing the whole population. Examples include
measurements of an evapotranspiration weather station monitoring a wind-
whipped hillside being equated to the rather calm conditions on the opposite
side, or the altered conditions and disturbed specimens in a lysimeter or
pressurized chamber to that of a heterogeneous field, or instantaneous remote
9
sensing every two weeks, or the frequent transfer of empirical methods from one
environment to another, or even the tendency in nature for climate to change.
A number of assumptions are usually applied to satisfy either the water or
energy balance to estimate water movements on the land surface (Equations 1.1
and 1.2). Presently, texture and nutrient classifications of soil layers in both the
vertical horizons and horizontal spread are interpolations of relatively few known
samples compared to the diversity of the landscape. Measurements of water
within the soil are therefore assigned values for porosity, specific yield, and
hydraulic conductivity on a broad basis that may not coincide with reality and
distort the actual flows and storage term in Equation 1.1.
Precipitation rates are also spatially variable, especially in mountainous
terrain, which may lead to inconsistencies when extrapolated from a few gaging
stations within a basin. This leaves the last component of the water balance, that
of losses returning to the atmosphere through evapotranspiration. Since
evapotranspiration cannot yet be adequately modeled from measurements of the
other variables at a high resolution, different approaches have been studied to
determine the amount of plant transpiration and surface evaporation.
ETQQPV OUTIN −−+=∆ [ 1.1 ]
Where V∆ is the change in system volumetric storage, P is precipitation,
INQ and OUTQ are the input and output volumes from the system, and ET is
evapotranspiration.
10
One more exclusively theoretical method in obtaining evapotranspiration
would be to solve for the latent heat flux in Equation 1.2 and then convert the
consumed energy to evaporated water. A positive aspect of this approach lies in
the relatively constant and known seasonal patterns of incoming solar radiation.
Determining the net radiation, however, including the surface albedo and
scattering and reflectance of light by particles in the atmosphere, varies not only
with position but also in time. Again, the composition and water content of the soil
greatly impacts its ability to intake and release heat while the sensible heat flux,
like the latent heat flux, is not directly measured. Further, this equation
addresses plant stored energy and processes such as photosynthesis in the
storage term which is frequently neglected altogether.
ETHGRE N λ−−−=∆ [ 1.2 ]
Where E∆ is the change in system energy storage, NR is net radiation, G
is the soil heat flux, H is the sensible heat flux, and ETλ is latent heat flux.
In order to determine evapotranspiration on a site and regional basis,
research from past decades has introduced empirical procedures into Equations
1.1 and 1.2 to account for lack of necessary environmental data.
1.2.1 Penman-Monteith Reference Evapotranspiration
H. L. Penman (1948) suggested a method for determining
evapotranspiration over natural surfaces by uniting both the energy balance
(Equation 1.2) and mass transfer equations to form what is now known as the
11
“combination equation”. This semi-empirical approach used observations of
sunshine, humidity, wind speed, and temperature in addition to site specific
conditions to estimate input parameters. With further research, Monteith (1965)
improved the theoretical and empirical relations and formed Equation 1.3. This
equation eventually proved superior to 19 similar approaches and became the
world standard (Allen et al. 1998). Inclusion in this study provides a means to
estimate evapotranspiration from readily available and complete datasets that
can be compared to more complex methodologies with incomplete datasets or
precluded output.
( ) ( )
+⋅+∆⋅
−⋅⋅+−⋅∆
=
A
S
A
ASPAN
RR
ReecGR
ET1γλ
ρ [ 1.3 ]
Where ET is evapotranspiration, ∆ is the slope of the saturation vapor
pressure-temperature curve, NR is net radiation, G is the soil heat flux, Aρ
is the dry air density, Pc is the specific heat capacity of air, AR is the
aerodynamic resistance, SR is the surface resistance, Se is the saturation
vapor pressure, Ae is the saturation vapor pressure, λ is the latent heat of
vaporization, and γ is the psychrometric constant.
In an effort to normalize assumptions and create an equation appropriate
for conditions in the United States, especially the arid West, an American Society
of Civil Engineers task force agreed upon the ASCE Standardized Penman-
12
Monteith equation or Equation 1.4 (Walter, 2000). As all of the study areas are
located within the western United States, this equation was selected as the most
appropriate and accepted Penman-combination method for this research.
( ) ( )( )2
2
1273
408.0
uC
eeuT
CGRET
d
ASn
N
SZ ⋅+⋅+∆
−⋅⋅+
⋅+−⋅∆⋅=
γ
γ [ 1.4 ]
Where T is temperature, 2u is wind speed, and nC and dC are reference
constants.
1.2.2 Bowen Ratio Energy Balance Technique
While both the solar and infrared radiative fluxes and soil heat flux can be
directly measured, the partitioning of the sensible and latent heat fluxes is much
more difficult to establish and must be determined indirectly. An example already
seen was the Penman-Monteith equation, which followed the energy balance
very closely with radiation and soil heat flux terms but quickly became convoluted
in evaluation of the sensible heat flux. As the sum of the sensible and latent heat
fluxes equals the difference between the radiative and soil heat fluxes,
discovering the ratio between the sensible and latent heat fluxes would enable
calculation of the fluxes. This method is called the Bowen ratio and is shown in
Equation 1.5. Since inputs of air temperature and vapor pressure at two different
vertical heights along with air pressure can be straightforwardly obtained, this
equation is easily implemented and maintainable. There are, however, limits to
its application as the theory is idealized to the case of turbulent atmosphere,
13
horizontal and homogeneous terrain, and negligible vertical advection, which are
not always the norm in the American West (Angus and Watts 1984, Perez et al.
1999).
eTcP
ETH P
∆⋅⋅∆⋅⋅
==ελλ
β [ 1.5 ]
Where β is the Bowen ratio, H is the sensible heat flux, ETλ is latent
heat flux, P is air pressure, Pc is the specific heat capacity of air, T∆ is
the change in temperature at two heights, λ is the latent heat of
vaporization, ε is the ratio of molecular weights of water to that of dry air,
and e∆ is the change in vapor pressure at two heights.
Describing behavior of the Bowen ratio numerically can be problematic as
it approaches asymptotes or changes sign in either the numerator or
denominator. An alternative but equivalent proportion is the fraction of latent
heat flux to that of the sum of both the latent and sensible heat fluxes. This
evaporative fraction ranges between 0 and 1 and is related to the Bowen ratio
through Equation 1.6.
β+=
11EF [ 1.6 ]
Where EF is the evaporative fraction and β is the Bowen ratio.
Once the evaporative fraction is calculated, it is multiplied against the net
available energy to produce evapotranspiration as is shown in Equation 1.7.
14
λGREFET N −
⋅= [ 1.7 ]
Where ET is evapotranspiration, EF is the evaporative fraction, NR is net
radiation, G is the soil heat flux, and λ is the latent heat of vaporization.
1.2.3 Eddy Covariance Technique
Some of the science behind the eddy covariance method involves parcels
of air (eddies) circulating through the atmosphere and along the earth’s surface
of diminishing sizes. These motions transfer particles, including suspended
water molecules, from one point in the air to another. Inverse of the frequency or
the period of the eddies is proportional to their size and may vary from less than
a second per cycle to many hours. Larger eddies containing most of the energy
ultimately shrink in size and dissipate energy at higher frequencies through what
is known as an energy cascade. By reducing the Navier-Stokes equation
through a number of assumptions similar to the Penman-Monteith and Bowen
ratio methods already mentioned to a homogeneous ground cover with sufficient
fetch, evapotranspiration is derived in Equation 1.8 as the covariance between
the vertical wind component and water vapor density (Aubinet et al. 2012).
'' VZuET ρ= [ 1.8 ]
Where ET is evapotranspiration, Zu is the vertical wind component, and
Vρ is water vapor density.
15
The advantage of this method over the Bowen ratio is the measurement of
advective flux; although, it is similarly handicapped by stable atmospheric
conditions. Sampling intervals of less than a second (10 to 20 Hertz) are
required to capture the covariance of the small eddies and must be calculated
over many minutes or hours to integrate the influence of the large eddies. This
leads to (for our current systems) large data files and processing requirements as
well as instrument spikes that must be corrected. Further adjustments due to
measurement practices may involve coordinate rotations, sensor lag times,
tapering or de-trending, frequency response correction, spectral corrections, the
krypton hygrometer oxygen correction, and the Webb-Pearman-Leuning (WPL)
correction (Lee et al. 2004). Periods of stable and neutral conditions must be
identified which may be during times of strong advective conditions or
questionable measurement. With the possibility for many unaccounted errors,
careful instrument setup and operation and data post processing will produce
acceptable results for time steps representative of the underlying assumptions.
1.2.4 Groundwater Fluxes
Plant water consumption patterns beneath the vadose zone are revealed
by measuring groundwater table fluctuations. Simple and accurate devices that
gauge depth such as pressure and temperature sensors, floats, or sonic meters
are capable of logging daily or finer readings. The difference in height from one
time step to another indicates the response in storage to the sum of
inflows/outflows and water extracted by plants, these physical interactions being
16
embodied in Equation 1.9. By definition, it excludes any evapotranspiration from
precipitation or irrigation that occurs in the top layer of unsaturated soil. Because
soil is a porous medium, nominal changes in height must be multiplied by the
ratio of the drainable volume to the whole (specific yield) to obtain actual
quantities. Complexity and irregularities in soil structure that can abruptly vary
from one horizon to the next have made it difficult to determine specific yield
beyond an educated range derived from disturbed samples or documented for
the prevalent texture(s). Problematic also is the rate at which recharge occurs
due to the unknown path through the soil and the peculiarities of the source.
Groundwater-based evapotranspiration can therefore only be estimated and not
measured unless the associating factors are known.
( )HRSET YG ∆−= * [ 1.9 ]
Where GET is groundwater evapotranspiration, YS is specific yield, R is
the change in depth due to lateral recharge or depletion, and H∆ is the
change in phreatic surface elevation.
Based on research completed in southwestern Utah, or dead-center of the
study area, White (1932) recommended estimating the recharge rate by
averaging the measured change in depth for the first four hours after midnight.
This assumes that no plant transpiration takes place during this time which may
not always be the case (Dawson et al. 2007). Another approach (Soylu et al.
2012) focuses on groundwater fluctuations caused by plant withdrawals during
the day and net recharge at night by removing linear and sinusoidal trends and
17
equating the recharge to the coefficient of the sinusoidal trend. Both of these
methods perform better averaged over multiple observations, but the recharge
could also be disaggregated and applied to the individual time steps.
1.2.5 Remote Sensing Energy Balance Approach
Remote sensing obviates a critical assumption upon which the previous
four methods are founded: that a point measurement adequately represents
surrounding conditions. Instead, distant instruments transmit and/or receive
electromagnetic waves and relate wave properties from distant sources to object
features such as surface temperature and location. Levels of spatial, temporal,
spectral, and radiometric resolution of the surveillance depend on the
instrument’s field-of-view, mobility, storage, communication, and design. While
the entire globe can be monitored by a single satellite in orbit, there is an
inherent tradeoff of either a loss in spatial or temporal coverage. Uncertainty in
the measurements may emanate from alterations in the signal’s path, additional
noise, and instrument calibrations. Nevertheless, remote sensing is a valuable
asset that can be used to estimate energy balance fluxes that can be verified by
point measurements.
Research in this field of evapotranspiration estimation using remote
sensing is ongoing, with several promising methodologies. One of them, the
Two-Source model, combines the energy balance at both the canopy and soil
surfaces as shown in Equation 1.10 (Kustas et al. 2005). With Landsat imagery
available (Vogelmann et al. 2001) for the study areas, intermittent estimates of
18
instantaneous evapotranspiration can be evaluated at the time of satellite
overpass and compared to other ground-based methods. This dataset would
afford analysis of surrounding areas that could explain differences in observed
behavior between instrument stations.
( ) ( )
+−
+−
⋅⋅−−+⋅=SA
AS
A
ACPNSNC RR
TTR
TTcGRRET ρλ1 [ 1.10 ]
Where ET is evapotranspiration, λ is the latent heat of vaporization, NCR
is net radiation over the crop, NSR is net radiation over the soil, G is the
soil heat flux, ρ is the air density, Pc is the specific heat capacity of air, CT
is the canopy temperature, ST is the soil temperature, AT is the air
temperature, AR is the aerodynamic resistance, and SR is the surface
resistance.
1.3 References
Allen, R.G. 1996. Assessing integrity of weather data for reference evapotranspiration estimation. Journal of Irrigation and Drainage Engineering, 122(2), 97-106.
Allen, R.G., Pereira, L.S., Raes, D., & Smith, M. 1998. FAO Irrigation and drainage paper No. 56. Rome: Food and Agriculture Organization of the United Nations, 26-40.
Angus, D. E., & Watts, P.J. 1984. Evapotranspiration—How good is the Bowen ratio method?. Agricultural Water Management, 8(1), 133-150.
Aubinet, M., Vesala, T., & Papale, D. (Eds.). 2012. Eddy covariance: a practical guide to measurement and data analysis. Springer.
Baum, B.R. 1978. The Genus Tamarix. Israel Academy of Sciences and Humanities, Jerusalem.
19
Blaney, H.F., & Criddle, W.D. 1962. Determining consumptive use and irrigation water requirements (No. 1275). US Department of Agriculture.
Christiansen, J.E. 1968. Pan evaporation and evapotranspiration from climatic data. Proc Amer Soc Civil Eng, J Irrig Drainage Div,, 94, 243-265.
Cleverly, J.R., Dahm, C.N., Thibault, J.R., Gilroy, D.J., & Allred Coonrod, J.E. 2002. Seasonal estimates of actual evapo-transpiration from Tamarix ramosissima stands using three-dimensional eddy covariance. Journal of Arid Environments, 52(2), 181-197.
CNWR. Accessed November 12, 2013. Cibola National Wildlife Refuge | Arizona and California. U.S. Fish & Wildlife Service. <http://www.fws.gov/refuge/Cibola/>
Dawson, T.E., Burgess, S.S., Tu, K.P., Oliveira, R.S., Santiago, L.S., Fisher, J.B., Simonin, K.A., & Ambrose, A.R. 2007. Nighttime transpiration in woody plants from contrasting ecosystems. Tree Physiology, 27(4), 561-575.
Di Tomaso, J.M. 1998. Impact, biology, and ecology of saltcedar (Tamarix spp.) in the southwestern United States. Weed technology, 326-336.
Geli, H.M., Neale, C.M., Watts, D., Osterberg, J., De Bruin, H.A., Kohsiek, W., Pack, R.T., & Hipps, L.E. 2012. Scintillometer-Based Estimates of Sensible Heat Flux Using Lidar-Derived Surface Roughness. Journal of Hydrometeorology, 13(4), 1317-1331.
Glenn, E.P., & Nagler, P.L. 2005. Comparative ecophysiology of Tamarix ramosissima and native trees in western US riparian zones. Journal of Arid Environments, 61(3), 419-446.
Gollehon, N., & Quinby, W. 2000. Irrigation in the American West: Area, water and economic activity. International journal of water resources development, 16(2), 187-195.
Han, W., Yang, Z., Di, L., & Mueller, R. 2012. CropScape: A Web service based application for exploring and disseminating US conterminous geospatial cropland data products for decision support. Computers and Electronics in Agriculture, 84, 111-123.
Hargreaves, G.L., Hargreaves, G.H., & Riley, J.P. 1985. Irrigation water requirements for Senegal river basin. Journal of Irrigation and Drainage Engineering, 111(3), 265-275.
Jensen, M.E. 1968. Water consumption by agricultural plants (Chapter 1).
20
Jensen, M.E., & Haise, H.R. 1963. Estimating evapotranspiration from solar radiation. Proceedings of the American Society of Civil Engineers, Journal of the Irrigation and Drainage Division, 89, 15-41.
Lee, X., Massman, W.J., & Law, B.E. (Eds.). 2004. Handbook of micrometeorology: a guide for surface flux measurement and analysis (Vol. 29). Springer.
Legates, D.R., & McCabe, G.J. 1999. Evaluating the use of “goodness-of-fit” measures in hydrologic and hydroclimatic model validation. Water Resources Research, 35(1), 233-241.
Li, F., Kustas, W.P., Prueger, J.H., Neale, C.M., & Jackson, T.J. 2005. Utility of remote sensing-based two-source energy balance model under low-and high-vegetation cover conditions. Journal of Hydrometeorology, 6(6), 878-891.
Makkink, G.F. 1957. Testing the Penman formula by means of lysimeters. J. Inst. Water Eng, 11(3), 277-288.
Mitchell, K.E., Lohmann, D., Houser, P.R., Wood, E.F., Schaake, J.C., Robock, A., Cosgrove, B.A., Sheffield, J., Duan, Q., Luo, L., Higgins, R.W., Pinker, R.T., Tarpley, J.D., Lettenmaier, D.P., Marshall, C.H., Entin, J.K., Pan, M., Shi, W., Koren, V., Meng, J., Ramsay, B.H., & Bailey, A.A. 2004. The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. Journal of Geophysical Research, 109(D7), D07S90.
Monteith, J.L. 1965. Evaporation and environment. In Symp. Soc. Exp. Biol (Vol. 19, No. 205-23, p. 4).
Nagler, P.L., Morino, K., Didan, K., Erker, J., Osterberg, J., Hultine, K.R., & Glenn, E.P. 2009. Wide-area estimates of saltcedar (Tamarix spp.) evapotranspiration on the lower Colorado River measured by heat balance and remote sensing methods. Ecohydrology, 2(1), 18-33.
NLDAS-2. The North American Land Data Assimilation System Phase 2. NASA Goddard Space Flight Center and Other Collaborators. <ftp://hydro1.sci.gsfc.nasa.gov/data/s4pa//NLDAS/NLDAS_FORA0125_H.002/>
Ozdogan, M., & Gutman, G. 2008. A new methodology to map irrigated areas using multi-temporal MODIS and ancillary data: An application example in the continental US. Remote Sensing of Environment, 112(9), 3520-3537.
21
Penman, H.L. 1948. Natural evaporation from open water, bare soil and grass. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 193(1032), 120-145.
Perez, P.J., Castellvi, F., Ibanez, M., & Rosell, J.I. 1999. Assessment of reliability of Bowen ratio method for partitioning fluxes. Agricultural and Forest Meteorology, 97(3), 141-150.
Priestley, C.H.B., & Taylor, R.J. 1972. On the assessment of surface heat flux and evaporation using large-scale parameters. Monthly weather review, 100(2), 81-92.
Pruitt, W.O., & Doorenbos, J. 1977. Empirical Calibration: A Requisite for Evapotranspiration Formulae Based on Daily Or Longer Mean Climate Data?. The Committee.
Shafroth, P.B., Cleverly, J.R., Dudley, T.L., Taylor, J.P., Riper III, C.V., Weeks, E.P., & Stuart, J.N. 2005. Control of Tamarix in the western United States: implications for water salvage, wildlife use, and riparian restoration. Environmental Management, 35(3), 231-246.
Sher, A., & Quigley, M.F. (Eds.). 2013. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press.
Soylu, M.E., Lenters, J.D., Istanbulluoglu, E., & Loheide, S.P. 2012. On evapotranspiration and shallow groundwater fluctuations: A Fourier-based improvement to the White method. Water Resources Research, 48(6).
Stromberg, J.C., Chew, M.K., Nagler, P.L., & Glenn, E.P. 2009. Changing perceptions of change: the role of scientists in Tamarix and river management. Restoration Ecology, 17(2), 177-186.
Taghvaeian, S. 2011. Water and Energy Balance of a Riparian and Agricultural Ecosystem along the Lower Colorado River. (Doctoral dissertation). All Graduate Plan B and other Reports. Paper 35.
Thornthwaite, C.W. 1947. Climate and moisture conservation. Annals of the Association of American Geographers, 37(2), 87-100.
Thornthwaite, C.W. 1948. An approach toward a rational classification of climate. Geographical review, 38(1), 55-94.
Turc, L. 1961. Estimation of irrigation water requirements, potential evapotranspiration: a simple climatic formula evolved up to date. Ann. Agron, 12(1), 13-49.
22
Vogelmann, J.E., Howard, S.M., Yang, L., Larson, C.R., Wylie, B.K., & Van Driel, N. 2001. Completion of the 1990s National Land Cover Data Set for the conterminous United States from Landsat Thematic Mapper data and ancillary data sources. Photogrammetric Engineering and Remote Sensing, 67(6).
Walter I.A., Allen R.G., Elliott R., Mecham B., Jensen M.E., Itenfisu D., Howell T.A., Snyder R., Brown P., Echings S., Spofford T., Hattendorf M., Cuenca R.H.,Wright J.L. & Martin D. 2000, November. ASCE’s standardized reference evapotranspiration equation. In Proc. of the Watershed Management 2000 Conference, June.
Westenburg, C.L., Harper, D.P., & DeMeo, G.A. 2006. Evapotranspiration by Phreatophytes Along the Lower Colorado River at Havasu National Wildlife Refuge, Arizona. US Geological Survey.
White, W.N. 1932. A method of estimating ground-water supplies based on discharge by plants and evaporation from soil: Results of investigations in Escalante Valley, Utah. US Government Printing Office.
23
CHAPTER 2*
COMPARISON OF THE NLDAS WEATHER FORCING MODEL TO
AGROMETEOROLOGICAL MEASUREMENTS IN THE WESTERN UNITED
STATES
2.1 Abstract
Diverse topography and climate in the American West have stymied
efforts to accurately quantify the flux of water on the land surface at a high spatial
resolution. Observations of the processes governing the earth’s water budget
and energy balance are generally from disparate point measurements on the
ground and have lower frequency, distinction, or confidence when remotely
sensed. Combined, these terrestrial and aerial sources can offset the other’s
inherent weaknesses. At a local scale, methodologies have been developed to
estimate evapotranspiration. A systematic approach to calculating reference
evapotranspiration at a regional scale over the western United States was
explored by comparing the drivers of the North American Land Data Assimilation
System weather forcing model to 704 agriculturally representative, electronic
weather stations at an hourly time step. Parameters of solar radiation, air
temperature, wind speed, and humidity were analyzed to identify any
*Reprinted from Journal of Hydrology, Elsevier, 10.1016/j.jhydrol.2013.12.040, (Volume 510, 14 March 2014, Pages 385–392). Clayton S. Lewis, Hatim M.E. Geli, Christopher M.U. Neale. “Comparison of the NLDAS Weather Forcing Model to Agrometeorological Measurements in the western United States”, Copyright (2014), Limited License for English Reproduction.
24
uncertainties and biases. Initial inspection of the weather parameter comparisons
revealed unsatisfactory performance of one or more of the NLDAS parameters in
several regions, but this was mollified in the calculation of reference
evapotranspiration for all but the southerly portions of the study area.
2.2 Introduction
While much effort has been expended by public and private institutions to
produce evapotranspiration (ET) estimates from inconsistent methodologies,
instrumentation, and temporal and spatial scales, a standard, invariable, and
accurate procedure to calculate reference ET from satellite and ground-based
sources on a regional basis is yet to be established. The resultant medley of
approaches evinces lack of coordination in an age of instant communication,
rapid data transfers, and high computing power. Conversely, the development of
spatial weather forcing data enhanced by remote sensing has offered increased
accuracy of ET estimates at the watershed level. As part of a US Geological
Survey undertaking, the goal of this study was to ascertain the fitness of a model
of assimilated and meteorological data in representing agricultural weather
conditions for reference ET calculations in the western United States. This was
achieved by comparing observations and calculated reference ET for weather
stations located in irrigated areas in the West to the model.
2.3 Background
NLDAS, or the North American Land Data Assimilation System, is a land
surface weather forcing model covering the conterminous United States at a
25
1/8th latitude-longitude degree grid. Support for NLDAS involved participants
from the National Oceanic and Atmospheric Administration (NOAA), the National
Aeronautics and Space Administration (NASA), Princeton University, the
University of Washington, the University of Maryland, and Rutgers University.
The intent was “to provide reliable initial land surface states to coupled
atmosphere-ocean-land models in an effort to improve weather predictions” (Xia
et al., 2012b). Weather parameters in NLDAS were downscaled from the North
American Regional Reanalysis (NARR)—which has a coarser grid in the Lambert
Conformal Conic projection with a 3 hour temporal resolution—and adjusted for
differences in elevation (Cosgrove et al., 2003b). NARR is modeled from surface,
airborne, and satellite datasets and NOAA’s National Centers for Environmental
Prediction Eta Model with lateral boundaries from the antecedent Global
Reanalysis (Mesinger et al., 2006). The product of interpolating NARR was an
hourly weather forcing dataset with a pixel size around 1/3rd of NARR centered
on the continental United States. NLDAS resolution increases with latitude, and
the pixel area ranges between 44-66 square miles (112-170 square kilometers),
which is too large to adequately depict the effects of mountainous terrain on
many land surface processes. Selected weather parameters for this study were
downward shortwave radiation, air temperature, wind speed, and derived relative
humidity (from surface pressure, air temperature, and specific humidity) at an
hourly time step from January 1, 1979 through December 31, 2012. These
weather parameters are the inputs needed to calculate reference ET with the
26
ASCE Standardized Penman-Monteith equation on an hourly time step (Walter et
al., 2000).
Initial comparison of NLDAS model output revealed small differences to
observed data around the state of Oklahoma (Luo et al., 2003). Robock et al.
(2003) evaluated four land surface models (Noah, Variable Infiltration Capacity,
Sacramento Soil Moisture Accounting, and Mosaic) during the warm season
forced by NLDAS for the same area and reported inconsistencies among the
land surface models and between those models and the observed data. Xia et al.
(2012a, 2012b) recounted other studies that exposed seasonal and spatial
anomalies, especially in arid and semi-arid regions. In the model spin-up for
NLDAS, Cosgrove et al. (2003a) acknowledged that for all simulations the
southwest region advanced the slowest and never reached satisfactory levels.
Those findings prompted major improvements to NLDAS, which was upgraded to
its second phase along with many adjustments to the land surface models.
Peters‐Lidard et al. (2011) compared reference ET from both versions of NLDAS
to flux towers and to a MODIS satellite algorithm and concluded that NLDAS
displayed a high bias and overestimated reference ET in the shoulder seasons
while ignoring the effects of irrigation. These all highlight shortcomings within
NLDAS in replicating land surface processes in the differential terrain, semi-arid
climates, and irrigated agriculture that characterize the Intermountain West.
The comparison dataset chosen to validate NLDAS in the West consisted
of agricultural weather station networks managed by university, state, and federal
organizations (Table 2.1). Many of the networks had similar equipment and data
27
collection practices, but none were identical in every aspect of instrumentation,
site considerations, data corrections, and data access. Periods of record ranged
from a few months to 32 years in the interval from 1981 to 2012 with a steady
increase in the number of stations reporting per year Fig. 2.1). Only networks or
stations logging at an hourly (or more frequent) time step with observations of
solar radiation, air temperature, wind speed, and humidity were selected. These
weather stations represented open, well-watered environments and
approximated reference conditions for estimating reference ET. Thus by
comparison of these observed datasets to NLDAS any warmer, drier, or calmer
biases in NLDAS, indicative of non-reference conditions, could be detected.
Fig. 2.1. Annual count of operational agricultural electronic weather stations in the western United States (1981-2012).
0
100
200
300
400
500
600
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011
Num
ber o
f Sta
tions
Year
28
Table 2.1. Electronic weather station networks over agricultural areas for 17 states in the western United States.
State Network Arizona AZMET California AgriMet, CIMIS, NICE Net Colorado AWDN, CoAgMet Idaho AgriMet Kansas WDL* Montana AgriMet, NDAWN Nebraska AWDN Nevada AgriMet, NICE Net New Mexico NM* North Dakota NDAWN Oklahoma Mesonet* Oregon AgriMet South Dakota NDAWN, SDAWDN Texas Texas ET, TXHPET Utah AgWxNet Washington AgriMet, AgWeatherNet Wyoming AgriMet, AWDN, WAWN All SCAN*
*Primary purpose not agricultural.
2.4 Data Preparation
2.4.1 Retrieval
Historical data from the agricultural weather station networks were
downloaded from the supervising agencies’ websites (see references for URLs)
excepting TXHPET and WAWN who both refused repeated requisitions. Due to
variable sampling rates, observations were aggregated and converted into hourly
total solar radiation, hourly average air temperature, hourly average wind speed,
29
and hourly average relative humidity. Assuming a logarithmic vertical profile,
wind speed measurements were standardized to a height of 6.6 feet (2 meters)
above the ground. Within the study area, electronic weather stations with
relevant parameters summed to 1134 by discounting 49 sites with essentially no
record.
Hourly raster datasets from the second phase of NLDAS (Forcing File A)
were downloaded from the NASA Goddard Space Flight Center repository
(NLDAS-2). Modeled weather parameters extracted from the GRIB file format
included surface downward shortwave radiation, air temperature, specific
humidity, surface pressure, and the u and v components of wind velocity.
Emulating the NLDAS budget bilinear interpolation procedure (Cosgrove et. al
2003b), weather variables were bilinearly interpolated to the electronic weather
station coordinates for the period of record at each site. Prior to interpolation,
solar radiation and air temperature were adjusted for differences in elevation
between the NLDAS pixels and the weather station site. A lapse rate of -18.83
Fahrenheit/mile (-6.5 Celsius/kilometer) was applied to the air temperature
values while the solar radiation data were multiplied by the ratio of the weather
station’s theoretical clear-sky solar radiation (from the ASCE Standardized
Penman-Monteith equation) to that of the pixel. Mitchell et al. (2004) reported
that NLDAS specific humidity was obtained by a terrain-height adjustment of the
air temperature and surface pressure while holding the relative humidity constant
from the NARR interpolation. Because relative humidity was common to the
agricultural networks, specific humidity in the NLDAS dataset was converted to
30
relative humidity to allow for a direct comparison. Calculation of the magnitude of
the wind velocity vectors preceded the interpolation, and the result was reduced
by a logarithmic wind height adjustment from 33 feet (10 meters) to a
standardized 6.6 feet (2 meters).
2.4.2 Weather Station Locale Filter
Processing procedures were developed to eliminate electronic weather
stations not representative of agricultural climatic conditions, of which all 16
weather networks contained sites not compliant. A systematic, fine-scaled
classification was sought to identify arid station sites. Allen (1996) described a
reference ET environment as a “uniform, well-watered…surface below and
surrounding the weather equipment for a distance of 100 times the height of the
wind, air temperature and relative humidity sensors in all directions.” Normally,
the sensors on the various platforms were positioned at a height from 5 feet (1.5
meters) to 10 feet (3 meters) possibly excluding the anemometer, the wind speed
nevertheless having been adjusted to a reference plane of 6.6 feet (2 meters).
According to Allen and given the height of 10 feet (3 meters), the watered fetch
about the station need only have extended 1000 feet (300 meters). Owing to the
imprecision of station coordinates documented by a few networks (e.g.,
coordinates rounded to the hundredth decimal of a latitudinal or longitudinal
degree [ten-thousandth of a radian]) the radial distance was increased to 0.5
miles (0.8 kilometers). Site suitability was determined by categorizing nearby
features within this buffer.
31
Next, existing methods to classify cultivated regions from non-agricultural
land use were reviewed. Ozdogan and Gutman (2008) mapped irrigated areas of
the continental United States using MODIS and other sources which appeared
accurate in the West but was limited to a spatial resolution of 0.3 miles (0.5
kilometers). Similar analysis has been conducted by the USDA National
Agricultural and Statistics Service (NASS) resulting in a 100 foot (30 meter)
resolution annual product of agricultural land use within the mainland United
States since 2010. Their approach processed images from several satellites,
surveys, land cover and elevation datasets, and other ancillary data to create a
detailed inventory entitled Cropland Data Layer (Han et al., 2012). One drawback
in the 100 plus crop types was the lack of discrimination between irrigated and
non-irrigated land. The fact that one-quarter of harvested cropland in the West is
irrigated (Gollehon and Quinby, 2000) would render the majority of a reference
ET environment classification non-irrigated if based solely on the agricultural
classes. However, the weather station sites that were removed were distant from
agriculture or located in urbanized zones, which could in either case be identified
by a nonexistent or low presence of cropland within a buffered radius. By
dissolving the NASS 2012 Cropland Data Layer into two classes of agricultural
and non-agricultural (Fig. 2.2), percent agriculture was calculated within the
buffered weather station circumference and a threshold of 30 percent designated
to infer reference ET conditions. 430 of the 1134 weather stations sites did not
meet this minimum requirement and were excluded from the analysis, the
balance of representative sites totaling 704.
32
Fig. 2.2. 704 agricultural electronic weather stations superimposed over the NASS 2012 Cropland Data Layer in the western United States.
Weather Station Measurement Filter Inherent to data collected from instruments outside of laboratory settings
are a multitude of potential errors that must be resolved prior to processing.
Errors evident in the weather station data included sensor malfunctions,
miscalibrations, spikes, power failures, station maintenance, datalogger program
33
changes, animal presence, covered pyranometers, frozen anemometers due to
ice or failed bearings, inconsistent units, duplicated observations, and
mishandled data storage. Of note was the combined weather station dataset
which summed to approximately 78 million observations each of hourly solar
radiation, air temperature, wind speed, and relative humidity and spanned 17
states and a variety of climates. Since manual screening of the data was
unfeasible, removal or correction, if possible, of the suspect records was
achieved through algorithms that relied on theoretical states and empirical
Boolean logic.
In light of the large quantity of available comparison data, emphasis was
placed on deletion over correction of suspect records. Independent subroutines
(detailed in Table 2.2) were used to examine each weather parameter to apply
unit conversions and calibrations to intervals of discontinuous data suggesting
sensor drift or drop-outs and nullify unrealistic, repeated, low, high, and sparse
measurements. Apparent was the level of maintenance, post-processing, and
care by each of the networks. Most were well-kept, a few mediocre, while the
worst performance in quality assurance in every category was the Texas ET
network of which three-quarters of the weather station sites were thrown out.
Altogether, 11.5 percent of the weather station data were rejected leaving an
average of 69 million observations per weather parameter or 0.77 percent of the
total NLDAS pixels located within the study area for the same time period.
34
Table 2.2. Sequential steps taken to programmatically filter and correct electronic weather station data.
On Download Unit conversions. Disaggregate running sums. Preprocessing Duplicate time steps overwritten in array. More frequent time steps converted to hourly averages or totals. Solar Radiation Set negative values to zero. Null repeated values for 24 or more hours. Null values below threshold for 240 or more hours. Null values where no more than 24 values are valid in 120 hours. Shift time to 0 UTC. Annually calibrate to clear-sky theoretical conditions. Air Temperature Null repeated values for 4 or more hours. Null improbable values. Null peaks (or dips) in values less than 5 degrees (0.087 radians). Null outliers outside a second-order polynomial fit and range. Find discontinuities within the data and repeat within sections: Check units. Null specks in data outside 2 standard deviations of a 96 hour interval. Null specks in data outside 2 standard deviations of a 960 hour interval. Null values where no more than 12 values are valid in 96 hours. Wind Speed Find discontinuities within the data and repeat within sections: Null improbable values. Null values below threshold for 120 or more hours. Null high 24 hour intervals. Null low 24 hour intervals. Null repeated values for 4 or more hours. Null peaks (or dips) in values less than 5 degrees (0.087 radians). Null values where no more than 48 values are valid in 120 hours.
35
Relative Humidity Null improbable values. Annually calibrate to quarterly mean maximum and minimum values. Null repeated values for 4 or more hours. Null peaks (or dips) in values less than 5 degrees (0.087 radians). Null excessively low 4 hour intervals. Null specks in data outside 3.5 standard deviations of a 48 hour interval. Null low 48 hour intervals. Null values where no more than 1 value is valid in 3 hours.
2.5 Comparison of NLDAS Data
2.5.1 Weather Parameters
Observations of solar radiation, air temperature, wind speed, and relative
humidity from the winnowed weather station datasets were contrasted against
NLDAS model interpolations spatially across the western United States
(examples shown in Fig. 2.3). Goodness of fit was assessed by considering the
coefficient of determination or R2, mean error of NLDAS from the weather station
measurement, and root mean squared error and interpolated for visualization
purposes only in Fig. 2.4, Fig. 2.5, and Fig. 2.6. Legates and McCabe (1999)
warned that validation of hydroclimatic models by correlation-based measures,
specifically R2, may result in erroneously high values when the model duplicates
extreme events. However, in the case of NLDAS, day-to-day fluctuations were
dampened while local variability was absorbed in the collective nature of the
NARR grid. In line with the comparison of Luo et al. (2003) of NLDAS in
Oklahoma, weather parameters agreed well, obviously calibrated to that state,
and degenerated in the more heterogeneous climate west of the prairie states.
36
Fig. 2.3. NLDAS and agricultural weather station scatter plots of hourly solar radiation, air temperature, wind speed, and relative humidity at four irrigated locations in the western United States.
Of all the weather parameters, modeled air temperature matched
measurements the best. High correlations of R2, frequently above 0.90, and low
errors characterized the plains east of the Rocky Mountains from North Dakota to
central Texas and the drier areas of the Pacific Northwest. Lower correlations
37
were observed along the coastal areas of the Pacific and in southern Texas,
albeit still surmounting a minimum R2 of 0.77. High biases of up to 7 Fahrenheit
(4 Celsius) degrees were seen in an annular shape from southern California into
Nevada, Utah, Idaho, and down through Wyoming, Colorado, New Mexico, and
Arizona with the error in the Four Corners area conversely low.
Solar radiation was the next best performing NLDAS parameter. R2
values ranged from 0.93 to 0.77 with portions of eastern Washington, eastern
Oregon, California, Arizona, and New Mexico out-performing the eastern plain
states. Strikingly, this same area also maintained the highest bias reaching 2.9
Langley per hour (34 Watt per square meter), which could have been related to
an error in the calibration process. Only Montana showed poorly in both
correlation and deviation with some question as to the accuracy of the
AgriMet(GP) network records. Noteworthy likewise were the low biases in
southern Texas with extremely high RMSE of 12 Langley (140 Watt per square
meter).
Relative humidity corresponded closely with air temperature, itself being a
function of temperature. However, values of R2 were significantly lower for
relative humidity than for air temperature, which may be remedied by correcting
the NLDAS air temperature bias before conversion. Densely irrigated yet semi-
arid zones in eastern Washington, central California, and northern Texas
correlated better with NLDAS relative humidity than areas with less widespread
irrigation, which could be attributed to similar non-arid datasets in the NLDAS
model.
38
Correlation of wind speed, like air temperature, was highest in the eastern
prairie states reaching a modest maximum of 0.72 but of deficient quality in
mountainous regions slumping to a low of 0.14 in California. Three strips along
the eastern, central, and western portions of the study area were biased high and
were interwoven with two low expanses. NLDAS underestimated wind speed in
the northern prairie states, severest in South Dakota, and overestimated in the
coastal areas of Texas and the Pacific Northwest.
2.5.2 Reference ET
Reference ET for a short crop height (grass) was estimated by the ASCE
Standardized Penman-Monteith equation, as aforementioned. Similar
comparisons of the R2, bias, and RMSE are shown in Fig. 2.7. Performance of
reference ET was surprisingly good with the lowest R2 surpassing any of the
individual input variables. Only in the four southern states (California, Arizona,
New Mexico, and Texas) and in an anomaly in central Kansas did NLDAS-forced
reference ET appear less reliable with respect to the bias. In lieu of further
analysis, the increased quality of reference ET as opposed to its drivers could be
attributed to the selective timing of evaporation and plant transpiration which
dissipates at night and lessens in the non-growing season. Further investigation
into NLDAS may highlight poorer fits in these intervals since the land surface
models for which it was tested may have behaved similarly to calculated
reference ET in this paper.
39
Fig. 2.4. IDW interpolated R Squared of hourly solar radiation, air temperature, wind speed, and relative humidity of the comparison between 704 agricultural weather stations and NLDAS in the western United States.
40
Fig. 2.5. IDW interpolated bias of hourly solar radiation, air temperature, wind speed, and relative humidity of the comparison between 704 agricultural weather stations and NLDAS in the western United States.
41
Fig. 2.6. IDW interpolated root mean squared error of hourly solar radiation, air temperature, wind speed, and relative humidity of the comparison between 704 agricultural weather stations and NLDAS in the western United States.
42
Fig. 2.7. IDW interpolated R Squared, bias, and root mean squared error of hourly short or grass reference evapotranspiration of the comparison between 704 agricultural weather stations and NLDAS in the western United States.
43
2.6 Conclusions
Validation of the second phase of the NLDAS weather-forcing model to
agrometeorological measurements in the western United States proved good for
parameters of air temperature and solar radiation, moderately fair for humidity,
and unacceptable for wind speed. Regions with little relief fared better than
mountainous regions even though the NLDAS grid was spatially refined to site
coordinates and elevation. More arid environments had higher temperatures and
lower humidity evidential of non-reference ET surroundings. Further, local
variations were drowned out by the aggregated nature of the low-resolution input.
These findings suggest that the temporal and spatial interpolation based on the
procedures outlined by Cosgrove et al. (2003b) do not adequately reproduce
local conditions (for which they were not intended) without resolving high and low
biases and seasonal and daily fluctuations. Although reflecting the biases,
calculation of reference ET seemingly bypassed the oscillating discrepancies. A
higher-resolution product is needed to account for irrigation, terrain differences,
and microclimates in order to estimate reference ET more accurately in the highly
diverse and semi-arid American West.
2.7 Acknowledgments
Tremendous appreciation is extended to all of the agricultural weather
station networks that maintain and make accessible a dataset applicable to
scientific research. The NLDAS data used in this study were acquired as part of
the mission of NASA’s Earth Science Division and archived and distributed by
44
the Goddard Earth Sciences (GES) Data and Information Services Center
(DISC). This study was funded in part by a US Geological Survey contract
(Award No. G11AP20229) and the Utah Agricultural Experiment Station. A
special thank you is also given to Mike Hobbins for his careful review of the
manuscript.
2.8 References
Allen, R.G. 1996. Assessing integrity of weather data for reference evapotranspiration estimation. Journal of Irrigation and Drainage Engineering, 122(2), 97-106.
Cosgrove, B.A., Lohmann, D., Mitchell, K.E., Houser, P.R., Wood, E.F., Schaake, J. C., Robock, A., Sheffield, J., Duan, Q., Luo, L., Higgins, R.W., Pinker, R.T., & Tarpley, J.D. 2003a. Land surface model spin-up behavior in the North American Land Data Assimilation System (NLDAS). Journal of Geophysical Research, 108(D22), 8845.
Cosgrove, B.A., Lohmann, D., Mitchell, K.E., Houser, P.R., Wood, E.F., Schaake, J.C., Robock, R., Marshall, C., Sheffield, J., Duan, Q., Luo, L., Higgins, R W., Pinker, R.T., Tarpley, J.D., & Meng, J. 2003b. Real-time and retrospective forcing in the North American Land Data Assimilation System (NLDAS) project. Journal of Geophysical Research, 108(D22), 8842.
Gollehon, N., & Quinby, W. 2000. Irrigation in the American West: Area, water and economic activity. International journal of water resources development, 16(2), 187-195.
Han, W., Yang, Z., Di, L., & Mueller, R. 2012. CropScape: A Web service based application for exploring and disseminating US conterminous geospatial cropland data products for decision support. Computers and Electronics in Agriculture, 84, 111-123.
Legates, D.R., & McCabe, G.J. 1999. Evaluating the use of “goodness‐of‐fit” measures in hydrologic and hydroclimatic model validation. Water Resources Research, 35(1), 233-241.
Luo, L., Robock, A., Mitchell, K.E., Houser, P.R., Wood, E.F., Schaake, J.C., Lohmann, D., Cosgrove, B., Wen, F., Sheffield, J., Duan, Q., Higgins, R. W., Pinker, R.T., & Tarpley, J.D. 2003. Validation of the North American
45
Land Data Assimilation System (NLDAS) retrospective forcing over the southern Great Plains. Journal of Geophysical Research, 108(D22), 8843.
Mesinger, F., DiMego, G., Kalnay, E., Mitchell, K., Shafran, P.C., Ebisuzaki, W., Jovic, D., Woollen, J., Rogers, E., Berbery, E.H., Ek, M.B., Fan, Y., Grumbine, R., Higgins, W., Li, H., Lin, Y., Manikin, G., Parrish, D., & Shi, W. 2006. North American regional reanalysis. Bulletin of the American Meteorological Society, 87(3), 343-360.
Mitchell, K.E., Lohmann, D., Houser, P.R., Wood, E.F., Schaake, J.C., Robock, A., Cosgrove, B.A., Sheffield, J., Duan, Q., Luo, L., Higgins, R.W., Pinker, R.T., Tarpley, J.D., Lettenmaier, D.P., Marshall, C.H., Entin, J.K., Pan, M., Shi, W., Koren, V., Meng, J., Ramsay, B.H., & Bailey, A.A. 2004. The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. Journal of Geophysical Research, 109(D7), D07S90.
Ozdogan, M., & Gutman, G. 2008. A new methodology to map irrigated areas using multi-temporal MODIS and ancillary data: An application example in the continental US. Remote Sensing of Environment, 112(9), 3520-3537.
Peters‐Lidard, C.D., Kumar, S.V., Mocko, D.M., & Tian, Y. 2011. Estimating evapotranspiration with land data assimilation systems. Hydrological Processes, 25(26), 3979-3992.
Robock, A., Luo, L., Wood, E.F., Wen, F., Mitchell, K.E., Houser, P.R., Schaake, J.C., Lohmann, D., Cosgrove, B., Sheffield, J., Duan, Q., Higgins, R.W., Pinker, R.T., Tarpley, J.D., Basara, J.B., & Crawford, K.C. 2003. Evaluation of the North American Land Data Assimilation System over the southern Great Plains during the warm season. Journal of Geophysical Research: Atmospheres (1984–2012), 108(D22).
Walter I.A., Allen R.G., Elliott R., Mecham B., Jensen M.E., Itenfisu D., Howell T.A., Snyder R., Brown P., Echings S., Spofford T., Hattendorf M., Cuenca R.H., Wright J.L. & Martin D. 2000, November. ASCE’s standardized reference evapotranspiration equation. In Proc. of the Watershed Management 2000 Conference, June.
Xia, Y., Ek, M., Wei, H., & Meng, J. 2012a. Comparative analysis of relationships between NLDAS‐2 forcings and model outputs. Hydrological Processes, 26(3), 467-474.
Xia, Y., Mitchell, K., Ek, M., Sheffield, J., Cosgrove, B., Wood, E., Luo, L., Alonge, C., Wei, H., Meng, J., Livneh, B., Lettenmaier, D., Koren, V., Duan, Q., Mo, K., Fan, Y., & Mocko, D. 2012b. Continental‐scale water
46
and energy flux analysis and validation for the North American Land Data Assimilation System project phase 2 (NLDAS‐2): 1. Intercomparison and application of model products. Journal of Geophysical Research: Atmospheres (1984–2012), 117(D3).
North American Land Data Assimilation System (Accessed May 2013) NLDAS-2. The North American Land Data Assimilation System Phase 2. NASA
Goddard Space Flight Center and Other Collaborators. <ftp://hydro1.sci.gsfc.nasa.gov/data/s4pa//NLDAS/NLDAS_FORA0125_H.002/>
Electronic Weather Station Networks (Accessed May 2013) AgriMet-GP. The Great Plains Cooperative Agricultural Weather Network.
Bureau of Reclamation: Great Plains Region, United States Department of the Interior. <http://www.usbr.gov/gp/agrimet/>
AgriMet-PN. The Pacific Northwest Cooperative Agricultural Weather Network. Bureau of Reclamation: Pacific Northwest Region, United States Department of the Interior. <http://www.usbr.gov/pn/agrimet/>
AgWeatherNet. The Washington Agricultural Weather Network. Washington State University Irrigated Agriculture Research and Extension Center. <http://weather.wsu.edu/awn.php>
AgWxNet. Utah Agricultural Weather Network. Utah Climate Center, Utah State University. <http://climate.usurf.usu.edu/agweather.php>
AWDN. Automated Weather Data Network. High Plains Regional Climate Center, University of Nebraska-Lincoln. <http://www.hprcc.unl.edu/awdn/>
CIMIS. California Irrigation Management Information System. California Department of Water Resources, University of California at Davis. <http://www.cimis.water.ca.gov/cimis/welcome.jsp>
CoAgMet. COlorado AGricultural Meteorological nETwork. Colorado Climate Center, Colorado State University. <http://www.coagmet.colostate.edu/>
Mesonet. Oklahoma Mesonet. Oklahoma Climatological Survey, University of Oklahoma. <http://www.mesonet.org/>
NDAWN. The North Dakota Agricultural Weather Network. Department of Soil Science, North Dakota State University. <http://ndawn.ndsu.nodak.edu/>
NICE Net. Nevada Integrated Climate and Evapotranspiration Network. Division of Hydrologic Science, Desert Research Institute. <http://nicenet.dri.edu/>
47
NM. NM Climate Center. New Mexico Climate Center, New Mexico State University. <http://weather.nmsu.edu/ws/network/nmcc/>
SCAN. Soil Climate Analysis Network. Natural Resources Conservation Service, United States Department of Agriculture. <http://www.wcc.nrcs.usda.gov/scan/>
SDAWDN. South Dakota Automated Weather Data Network. South Dakota Office of Climatology, South Dakota State University. <http://climate.sdstate.edu/climate_site/ag_data.htm>
Texas ET. Texas ET Network. Texas A&M AgriLife Extension Service. <http://texaset.tamu.edu/>
TXHPET. Texas High , North, and South Plains ET Networks. Texas A&M AgriLife Research and Extension, Amarillo and Lubbock. < https://txhighplainset.tamu.edu/login.jsp>
WAWN. Wyoming Agricultural Weather Network. Department of Plant Sciences, University of Wyoming. <http://www.wawn.net/>
WDL. Weather Data Library. Kansas State University Research and Extension. <http://wdl.agron.ksu.edu/>
48
CHAPTER 3 BOWEN RATIO AND EDDY COVARIANCE SYSTEM PERFORMANCE OVER
SALTCEDAR UNDER DIFFERENT ATMOSPHERIC STABILITY AND
ADVECTION CONDITIONS
3.1 Abstract
Sonoran Desert winds near the Colorado River bordering Arizona and
California increase the potential for evapotranspiration of riparian vegetation,
including the invasive and predominant saltcedar (Tamarix ramosissima). In
2006, energy balance flux towers were positioned in the floodplain over dense
stands of saltcedar to monitor the surface energy fluxes using both eddy
covariance and Bowen ratio energy balance techniques. After a fire destroyed
both the saltcedar forest and instruments in 2011, the towers were reconstructed
and measurements resumed. Influence of sensible heat advection was
examined by comparing the sensible and latent heat fluxes measured by the
more direct eddy covariance approach to the gradient Bowen ratio technique.
The turbulent transfer coefficient for heat was found to not equate to water vapor
throughout the day, but rather as a fraction that decreased linearly at one
location pre fire and remained constant at another location post fire.
3.2 Introduction
Saltcedar species (genus Tamarix), an invasive phreatophyte in the
western hemisphere transported from Eurasia in the nineteenth century A.D.,
have been the subject of numerous ecological research studies over the past six
49
decades due to their capability to outcompete native vegetation (such as willows
and cottonwoods [genera Salix and Populus]). This shrub or tree can grow to the
height of 5-12 meters and, upon leaf fall, deposits salts accumulated on its tiny
leaves on the soil surface. Saltcedar is susceptible to fire, flooding stream
channels, and propagation through high seed production, adventitious roots, and
root crown regrowth. Being a facultative halophyte combined with its tolerance of
flooding, drought, temperature extremes, cutting, and quick reestablishment after
a fire, makes saltcedar a formidable invasive competitor. Reported excessive
consumption of groundwater has targeted saltcedar for removal, markedly in
areas that can ill afford the loss such as the arid and semi-arid portions of the
American Southwest. Estimation of evapotranspiration (ET) rates have been
calculated from diverse methodologies, including lysimeters, groundwater
fluctuations, catchment stream flow balances, sap flow techniques, transpiration
chambers, scintillometers, Bowen ratio energy balance (BREB), eddy covariance
(EC), and remote sensing. (Di Tomaso, 1998; Shafroth et al., 2005) Each of
these techniques suffers from its own assumptions and limitations whether it be
spatial heterogeneity, fetch requirements, atmospheric stability, measurement
ranges and errors, or reliance on and behavior of unknown physical properties.
For systems modeling only the vertical transfer of energy, increased ET caused
by horizontal advection of sensible heat from hotter and drier localities to cooler
and wetter oases present a plausible source of underestimation of ET in
saltcedar stands located along riparian corridors in the deserts of the
southwestern United States. Within a region infested with saltcedar along the
50
Lower Colorado River in the United States, the role of heat advection from
adjacent desert lands was studied using both BREB and EC techniques.
Particular to this paper is the comparison of results both before and after a
wildfire charred the saltcedar and blackened the landscape.
3.3 Theoretical Foundations
Radiative energy from the sun warms both the earth’s atmosphere and is
the principal driver of evaporation. Near the surface, radiation is reflected,
emitted, consumed in processes of photosynthesis and ET, converted to heat, or
stored. By neglecting the normally minor components, this phenomenon can be
modeled by Equation 3.1 as
0=−−− EHGRN λ [ 3.1 ]
Where NR is the net flux of shortwave and longwave radiation, G is the soil
heat flux,H is the sensible heat flux, and Eλ is the latent heat flux.
While it is possible to directly and independently measure all the surface
fluxes to some degree, net radiation and the soil heat flux are more easily
measured or estimated through geometrical relationships, empirical calibrations,
soil properties, and long-term averages. By determining the energy applied to
sensible heat, the remainder may be attributed to latent heat and vice versa.
Another angle would be to determine the ratio of the two and then apportion
afterwards the remaining energy calculated from Equation 3.1. This is known as
the Bowen ratio, with the following definition:
51
EHλ
β = [ 3.2 ]
where β is the Bowen ratio.
For the EC method, the sensible and latent heat fluxes are directly
calculated from high frequency wind vector and atmospheric variable
measurements, which also includes numerous corrections (Aubinet et al., 2012).
BREB Bowen ratio estimates are found by measuring air temperature and water
vapor pressure offset vertically at two heights and by assuming a gradient of
those atmospheric properties between the two positions (Perez et al., 1999).
eT
kkPC A
V
HPGRADIENT δ
δλε
β = [ 3.3 ]
where GRADIENTβ is the Bowen ratio derived from the BREB method, PC is
the specific heat of air, P is the air pressure, Hk is the turbulent or eddy
transfer coefficient for heat, AT is the air temperature, λ is the latent heat of
vaporization or sublimation of water, ε is the molecular ratio of water
vapor to air, Vk is the turbulent or eddy transfer coefficient for water vapor,
and e is the vapor pressure of water.
In practice, a fundamental simplification to Equation 3.3 is infused by
assuming the eddy transfer coefficients for water vapor and heat are equal and
hence eliminated. However, the theoretical ideal is not always upheld, and
means to check calculations is wanting. By comparing the Bowen ratios
52
obtained from both EC and BREB methodologies, the value of the ratio of the
turbulent transfer coefficients may be back calculated (Papaioannou et al., 1989).
This would assume that EC was the standard and that any difference in Bowen
ratios were due to incongruent turbulent transfer coefficients.
GRADIENT
FLUX
V
H
kk
ββ
≅ [ 3.4 ]
where FLUXβ is the Bowen ratio derived from the EC method fluxes and
Equation 3.2.
One other useful equation when working with the Bowen ratio is the
conversion to the evaporative fraction, or the fraction of latent energy to net
radiation minus the soil heat flux, and can be obtained from Equations 3.1 and
3.2 as
β+=
11EF [ 3.5 ]
where EF is the evaporative fraction.
3.4 Methodology Limitations
Idealized states, upon which EC and BREB are established, frequently do
not represent measurement site conditions. Further, the extent of the theory may
not always be satisfied by predetermined assumptions. As an example, Alfieri et
al. (2012) found that variations in surface conditions could add uncertainty into
flux measurements. These variations could lead to unbalanced energy entering
into and exiting a measurement area and not fulfill conservation of energy as a
53
closed system. For EC, “surface fluxes can be under-measured for a number of
reasons including mismatched sources of [latent and sensible heat],
inhomogeneous surface cover and soil characteristics, flux divergence or
dispersion, non-stationarity of the flow, lack of a fully developed turbulent surface
layer, flow distortion, sensor separation, topography and instrument error” (Twine
et al., 2000). These deviations can negate the statistical requirements of EC
where time averages equate to ensemble averages, thus nullifying the ergodic
hypothesis (Aubinet et al., 2012). Other disparities can be caused by a
dynamically changing flux footprint with the wind vector, which is compounded by
intervals of low turbulence and the inability to properly account for advection
(Wilson et al., 2001). Under strongly advective periods, Alfieri et al. (2012)
emphasize the discrepancy of EC measurements.
Beyond standard sensor corrections, errors in the surface energy balance
[Equation 3.1] for EC are related to low frequency fluxes, advection, and
heterogeneous terrain (Aubinet et al., 2012). In the literature, there have been
two procedures proposed to force closure: the first, to assume the remaining
energy is all unaccounted for latent heat, and the second, to partition the residual
between latent and sensible heat in proportion to the Bowen ratio. Brotzge and
Crawford (2003), in a study comparing EC and BREB, found that the lowest
closure rates for EC corresponded with a lower latent heat flux when compared
to BREB therefore strengthening the position of the first option for forcing
closure. However, many authors strongly recommend the second approach as
54
the most suitable and logical (Twine et al., 2000; Chávez et al., 2009; Aubinet et
al., 2012).
Any environmental deviation from vertically transferred, gradient, and
closely associated sensible and latent fluxes produces uncertainty in BREB
measurements (Perez et al., 1999). Horizontal fluxes must be assumed
negligible (Figuerola and Berliner, 2005), which does not characterize regions of
advection. Also, very dry conditions—when the Bowen ratio is large and
positive—exceed the accuracy of BREB equipment and increasingly amplify the
error (Angus and Watts, 1984; Fuchs and Tanner, 1970). Adequate siting of a
BREB measurement station, based on the foregoing criteria, can then be
described as a flat, well-watered, and homogeneous landscape with sufficient
fetch in every direction (Brutsaert, 1982). Heilman et al. (1989) determined from
their measurements that BREB was insensitive to poor fetch at low Bowen ratios
and required a fetch-to-instrument-height ratio of 20:1, significantly lower than the
standard practice of 100:1. When imperfectly sited, BREB measurements can be
compared to a more direct method, such as lysimeters or EC, to determine the
accuracy (Perez et al., 1999).
Some errors are inherent to BREB, which becomes unwieldy when β is
near minus one and the evaporative fraction’s denominator [Equation 3.5] goes
to zero. Tanner et al. (1987) recommended restriction to the following domain to
omit this effect.
}75.025.1|{ −≥∪−≤ βββGRADIENT [ 3.6 ]
55
Additional instrument offset errors can be eliminated by swapping the
sensors on a periodic basis between measurement heights (Fuchs and Tanner,
1970). Payero et al. (2003) presented a procedure for identifying and removing
invalid data within BREB as a function of the direction of the fluxes:
( ) 0<−
+ GRTe
Nδλδ [ 3.7 ]
Another source of error is the hypothesis that the turbulent transfer
coefficients of heat and water vapor in Equation 3.3 are equal and whose effect
may be neglected (Gavilán and Berengena, 2007). This is true for neutral
atmospheric conditions, but may not fully represent unstable or stable conditions
with significant advection of the fluxes (Motha et al., 1979; Perez et al., 1999).
Within a stable internal boundary layer, the ratio of the exchange coefficients is
most often greater than unity when the Bowen ratio is less than zero and less
than unity when the Bowen ratio is greater than zero — with opposite behavior
experienced in an unstable internal boundary layer (Bink, 1997; Lee et al., 2004).
This means that under sensible heat advection (negative Bowen ratio) and
turbulent conditions (unstable), the coefficient ratio should be less than one
resulting in an underestimation of latent heat if ignored. Even then, Todd et al.
(2000) suggested that the interaction of the exchange coefficients under those
conditions is uncertain. Under advection, flux profiles drastically alter with heat
and water vapor being transported in opposite directions (Verma et al., 1977).
56
3.5 Advection and Atmospheric Stability
Advection occurs when mass or energy is carried in a fluid from one
location to another where the property is no longer in equilibrium (Baldocchi et
al., 2000b). When dealing with near earth fluxes, this may be caused by
heterogeneity in surface roughness due to changes in height, moisture, or
temperature (Alfieri et al., 2012). Horizontal sensible heat advection ensues
when relatively hotter air is transported into a more humid and thus cooler
environment. This imported thermal energy increases the vertical latent heat flux
over the surface by subtracting from the sensible heat flux in Equation 3.1.
Regional advection is defined as a constant addition to latent heat flux above that
supplied by net radiation over an area of interest, whereas local advection
distinguishes the waning contribution of sensible heat due to a sudden surface
transition up to an asymptotic or equilibrium level some distance into a fetch
(Brakke et al., 1978; Lee et al., 2004).
BREB temperature and humidity measurements at two heights must be
located within the internal boundary layer but also high enough to account for
heterogeneity at the surface. Under advective conditions, the gradient layer does
not exist and may introduce errors in flux measurements since the evaporation
rate at the surface exceeds that at a higher level to satisfy conservation of vapor
(Monteith 1965, Baldocchi et al., 2000b). When horizontal fluxes are
inconsistent, profiles and accompanying exchange coefficients are dissimilar and
not equal, thus destroying the BREB hypothesis (Angus and Watts, 1984). Todd
et al. (2000) reported that two BREB stations with sensors at the same height
57
and location agreed closely but that changing sensor heights increased
differences. Horizontal advection in BREB measurements can be spotted when
the mean daily surface flux of vertical sensible heat is negative or through a
weaker correlation with friction velocity (Berengena and Gavilán, 2005; van
Gorsel et al. 2008).
Fluctuations in atmospheric stability, driven by diurnal and seasonal cycles
of atmospheric-surface energy interactions, affect the influence and severity of
advection (Guo and Schuepp, 1994). Under very stable conditions and little
turbulence, advection can be the principal form of energy transfer in contrast to
more frequent conditions where it takes a lesser role (Baldocchi et al., 2000a).
With only small changes in atmospheric stability, large variations in exchange
coefficients can occur (Mahrt and Ek, 1984). Additional effects include the
development of flux footprints at different observation levels yielding inconstant
Bowen ratios with height (Leclerc and Thurtell, 1990; Figuerola and Berliner,
2005). In all, advection is prevalent in arid and semiarid areas, where sensible
heat greatly enhances the latent heat flux or ET in riparian and irrigated zones
and distorts flux measurements (Berengena and Gavilán, 2005).
3.6 Site Characteristics
Encompassing 7,464 hectares along the lower Colorado River, the Cibola
National Wildlife Refuge (Cibola) is maintained by the U.S. Fish and Wildlife
Service as a habitat for migratory birds but also affords opportunities for hunting,
fishing, wildlife watching, recreation, and environmental education and research
58
(USFWS-CNWR 2014). It is located within the Sonoran Desert, exhibits high
temperatures and little precipitation, and straddles the state lines of California on
the west and Arizona on the east (Fig. 3.1). Near the shoreline of the Colorado
River are stands of native cottonwood and willow, but the floodplain has been
overrun with invasive saltcedar (Tamarix ramosissima) interspersed with native
arrowweed. With interest in quantifying saltcedar transpiration as fed by shallow
groundwater in Cibola, several universities and government agencies have
installed and monitored groundwater depth loggers, EC and BREB systems,
sctintillometers, and sap flux instruments beginning in 2006. A number of
publications have documented the groundwater fluctuations, BREB,
scintillometers, and sap flux and have been summarized recently by Glenn et al.
(2013). What has not been previously analyzed is the EC data, a more direct
approach but more difficult to calculate than BREB. Given the extremes of the
barren wasteland that surround Cibola, the likelihood of advection inducing error
in BREB measurements is high. The purpose of this study was to compare the
surface fluxes obtained from EC to validate BREB under advection, from which
numerous results have already been published.
Atmospheric measurements were recorded at three measurement towers,
depicted in Fig. 3.1, overlooking the saltcedar forest denominated Slithern,
Diablo, and Swamp. Positioning of the towers resulted in three fetch and
groundwater conditions: next to the river (Swamp), full saltcedar growth
(Slithern), and medium saltcedar growth (Diablo). Originally, all three towers had
BREB systems, but only Diablo maintained EC. In August 2011, a wildfire
59
ravaged the Refuge and decimated the saltcedar along with the towers (Fig. 3.2).
Two of the towers were rebuilt in 2012, Slithern and Diablo, and were both
equipped with EC but only Slithern was complemented with BREB. This
arrangement created two datasets for comparison between EC and BREB: at
Diablo before the fire and afterwards at Slithern measuring the regrowth.
Fig. 3.1. Location of three surface flux measurement sites over a salt cedar forest within the Cibola Wildlife National Refuge near the Colorado River.
60
Fig. 3.2. Measurement towers and surrounding salt cedar forest pre and post fire at the Cibola National Wildlife Refuge [(A) Diablo tower pre fire, 2008 (B) Slithern tower pre fire, 2008 (C) Slithern tower after fire, 2012 (D) rebuilt Slithern tower post fire, 2013].
3.7 Data Analysis
Because of the remoteness of the study site, access to the measurement
platforms and maintenance by personnel was usually limited to the saltcedar
active growing season at an intermittence of around six weeks. Raw datasets
were then collected, inspected, stored, and backed up. In consequence of the
intemperate environment were a sensor or power source to fail, gaps in the data
record would follow. This was inevitably experienced, albeit the lesser portion of
the dataset. Dataloggers and EC instruments were purchased from Campbell
Scientific, Inc. (Logan, Utah) and comprised the CSAT3 3D sonic anemometer
61
and KH20 krypton hygrometer while the BREB equipment was acquired from
Radiation and Energy Balance Systems, Inc. (Seattle, Washington).
Latent and sensible heat fluxes were calculated from the EC data using
software customized for the project. No filters were applied to the data, but
spikes for each weather parameter were removed iteratively by exceeding a
multiple of the standard deviation from the calculation period mean. From ogive
testing, an hourly calculation interval was confirmed to constitute the bulk of the
low frequency eddies derived from the 10 Hz input. Several corrections (covered
in detail by Aubinet et al., 2012), which consisted of time shifting, traditional
coordinate rotation, frequency response, high frequency, spectral, KH20 oxygen,
and WPL were conducted. Net radiation and soil heat flux were logged on a 30
minute time step (from more frequent sampling) but were downgraded
(averaged) to an hour to force closure on the sensible and latent heat fluxes.
Since the second closure option using the Bowen ratio was chosen, the Bowen
ratio was preserved.
Next, BREB surface fluxes were considered. On the tower, the
measurement arms were mounted on a vertical exchange device that rotated the
top and bottom sensors every 15 minutes. These observations were averaged to
an hour to remove the sensor offset bias and align with the EC output. Any
aggregated time steps that contained invalid readings were immediately
discarded. From the input dataset at each location, the Bowen ratio was
calculated [Equation 3.3] and was filtered by Equations 3.6 and 3.7, if applicable.
62
Latent and sensible heat flux for each hour was also output in accordance with
the calculated Bowen ratio.
Once the flux and gradient (EC and BREB) Bowen ratios were calculated,
the outputs were adjusted for the overlapping period of record at each tower. For
Diablo, an EC system was present for the years 2008-2009 and 2012-2013, but
the BREB system was destroyed in the fire so only the first portion could be
compared. At Slithern, EC data was collected in 2012-2013 with its BREB
system also expiring in the flames, but, fortunately, the BREB system was
replaced with instruments that had been used in alfalfa fields earlier in the
project. These contiguous periods at Diablo pre and Slithern post fire were
evaluated in Equation 3.4 to determine the relationship between the eddy
transfer coefficients of heat and water vapor.
3.8 Results and Discussion
Flux and gradient Bowen ratios were compared for corresponding hours at
Diablo and Slithern Towers and are plotted in Fig. 3.3 and Fig. 3.4, respectively.
Positive Bowen ratios agree linearly between the two systems but are high,
meaning BREB apportioned more energy to sensible heat than EC did. It would
be apparent from the scatter of negative Bowen ratios between EC and BREB
that BREB completely failed to account for advection, but these values almost
exclusively occurred during the nighttime hours where stable conditions
undermined the fundamental assumptions of both methodologies. These
behaviors are congruent for both locations, before and after the fire. For clarity,
63
Fig. 3.3. Hourly fraction of Bowen ratios obtained from eddy covariance and Bowen ratio energy balance instrumentation over salt cedar at Diablo station in Cibola, California, pre fire (2008-2009).
the horizontal band with no data on the gradient axis of each graph is due to the
elimination of problematic Bowen ratios as per Equation 3.6.
By dividing the EC Bowen ratios by those obtained from BREB, the
proportional relationship between the heat and water vapor turbulent transfer
coefficients was determined. These were averaged at each hourly interval and
revealed telling trends as shown in Fig. 3.5. Diablo indicates unity between the
exchange coefficients in the early morning, but the heat profile then steadily
decreases throughout the day. Such occurrence could be attributed to its closer
proximity to the desert’s edge with less dense growth to compensate for
-10
0
10
-10 0 10
β GR
AD
IEN
T (U
nitle
ss)
βFLUX (Unitless)
64
Fig. 3.4. Hourly fraction of Bowen ratios obtained from eddy covariance and Bowen ratio energy balance instrumentation over salt cedar at Slithern station in Cibola, California, post fire (2012-2013).
increasing advection throughout the day. This is contrasted with Slithern, post
fire, where the correlation is held constant at around 0.6. For both locations,
BREB underestimates latent heat flux day and night; although nighttime fluxes
have a minor bearing on the overall energy transfer.
Hourly surface fluxes were averaged at Diablo and Slithern and the
averaged hourly relationship between the turbulent transfer coefficients
subtracted to correct the latent and sensible heat fluxes (Fig. 3.6 and Fig. 3.7).
These results illustrate the difference between how EC partitioned the fluxes as
opposed to BREB. At Diablo, the uncorrected BREB latent heat flux declined
-10
0
10
-10 0 10
β GR
AD
IEN
T (U
nitle
ss)
βFLUX (Unitless)
65
throughout the day suggesting reduced saltcedar stomatal response to elevated
saturation deficit, but when the correction was included the latent and sensible
heat fluxes evened out in the later hours. This behavior diverges from what
Nagler et al. (2009) described of sap flux measurements at Diablo Tower in
2008, which more closely adhered to the BREB initial results. Latent heat flux at
Slithern increased throughout the day but kept the same curvature, lining up with
pre-fire sap flux measurements. At both sites and for both the uncorrected and
corrected values, essentially no nighttime ET was measured, which also is in
contrast to the previous authors’ results that indicate a sizeable percentage of the
total daily ET occurring at night according to the sap flux.
Fig. 3.5. Average hourly fraction of Bowen ratios obtained from eddy covariance and Bowen ratio energy balance instrumentation over salt cedar in Cibola, California, Pre and post fire (2008-2013).
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
β FLU
X/β G
RA
DIE
NT
(Uni
tless
)
Hour Diablo (2008-2009) Slithern (2012-2013)
66
Fig. 3.6. Average hourly surface fluxes obtained from Bowen ratio energy balance instrumentation over salt cedar at Diablo station in Cibola, California, pre fire (2008-2009).
Fig. 3.7. Average hourly surface fluxes obtained from Bowen ratio energy balance instrumentation over salt cedar at Slithern station in Cibola, California, post fire (2012-2013).
-200
-100
0
100
200
300
400
500
600
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
(W/m
2 )
Hour Rn G H H_Corrected λE λE_Corrected
-200
-100
0
100
200
300
400
500
600
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
(W/m
2 )
Hour Rn G H H_Corrected λE λE_Corrected
67
For either measurement tower, BREB underestimated latent energy
consumption when compared to EC. Mean period fractions of the corrected and
uncorrected fluxes (listed in Table 3.1) imply that ET is estimated at 76 and 83
percent of assumed levels for Diablo and Slithern. Advection of sensible heat
flux from the surrounding Sonoran Desert is the most probable cause. Variations
in the hourly eddy transfer coefficient fraction indicate that the advection is more
of a local nature at Diablo and regional at Slithern. Of course, it would have been
helpful to compare both locations before and after the fire where canopy height
changed from around 6 meters to up to 2 meters of regrowth. This would have
distinguished the fire’s impact on the analysis, ascribing the remaining
divergence to the locality. A difference between the two sites is the tree density
that could be a function of distance from the river and groundwater salinity,
where Diablo reads twice the electrical conductivities of Slithern.
Table 3.1. Advection-corrected to non-corrected fraction of seasonal sensible and latent heat fluxes over salt cedar in Cibola, California, pre and post fire.
Station Period H λE Diablo 2008-2009 0.81 1.63
Slithern 2012-2013 0.58 1.80
68
3.9 Summary and Conclusions
Hourly estimated latent and sensible heat fluxes above saltcedar
vegetation were obtained from adjacent EC and BREB systems installed on two
separate towers in the Cibola National Wildlife Refuge, California. With the arid
environment surrounding the Colorado River, the suspicion was that sensible
heat advection would negatively affect the BREB partition of the latent heat flux.
EC and BREB datasets were prepared and used for comparison of calculated
surface fluxes. It was found that the BREB system at both locations linearly
overestimated the Bowen ratio and consequently underestimated the latent heat
flux during turbulent, daytime periods and was of no confidence at night due to
both systems being unable to measure accurately under stable conditions.
Inferred from the EC and BREB Bowen ratios, the fraction of the turbulent
transfer coefficients of heat and water vapor in the BREB Bowen ratio calculation
was shown to decrease during the day at one location and remain constant at
another, both well below unity. These lower exchange coefficients for heat
substantiated advective conditions at both towers under gradient BREB
methodology. In summary, uncorrected fluxes were found to underestimate and
misrepresent hourly contributions of the latent heat flux, or saltcedar ET.
3.10 Acknowledgments
This study was supported by the Six Agency Committee of the Colorado
River and the U.S. Bureau of Reclamation through a cooperative agreement.
Partial support was also provided by the Utah Agricultural Experiment Station
69
and the Remote Sensing Services Laboratory in the Department of Civil and
Environmental Engineering at USU. Appreciation is extended to the unnamed
field technicians and reviewers who contributed to this work.
3.11 References
Alfieri, J.G., Kustas, W.P., Prueger, J.H., Hipps, L.E., Evett, S.R., Basara, J.B., Neale, C.M., French, A.N., Colaizzi, P., Agam, N., Cosh, M.H., Chavez, J.L., & Howell, T.A. 2012. On the discrepancy between eddy covariance and lysimetry-based surface flux measurements under strongly advective conditions. Advances in Water Resources, 50, 62-78.
Angus, D.E., & Watts, P.J. 1984. Evapotranspiration—How good is the Bowen ratio method?. Agricultural Water Management, 8(1), 133-150.
Aubinet, M., Vesala, T., & Papale, D. (Eds.). 2012. Eddy covariance: a practical guide to measurement and data analysis. Springer.
Baldocchi, D.D., Meyers, T.P., & Wilson, K.B. 2000a. Correction of eddy-covariance measurements incorporating both advective effects and density fluxes. Boundary-Layer Meteorology, 97(3), 487-511.
Baldocchi, D.D., Finnigan, J., Wilson, K., Paw U, K.T. & Falge, E. 2000b. On measuring net ecosystem carbon exchange over tall vegetation on complex terrain. Boundary-Layer Meteorology, 96(1-2), 257-291.
Berengena, J., & Gavilán, P. 2005. Reference evapotranspiration estimation in a highly advective semiarid environment. Journal of irrigation and drainage engineering, 131(2), 147-163.
Bink, N.J. 1997. The ratio of eddy diffusivities for heat and water vapour under conditions of local advection. Physics and Chemistry of the Earth, 21(3), 119-122.
Brakke, T.W., Verma, S.B., & Rosenberg, N.J. 1978. Local and regional components of sensible heat advection. Journal of Applied meteorology, 17(7), 955-963.
Brotzge, J.A., & Crawford, K.C. 2003. Examination of the surface energy budget: A comparison of eddy correlation and Bowen ratio measurement systems. Journal of Hydrometeorology, 4(2).
70
Brutsaert, W. 1982. Evaporation into the atmosphere: theory, history, and applications.
Chávez, J.L., Howell, T.A., & Copeland, K.S. 2009. Evaluating eddy covariance cotton ET measurements in an advective environment with large weighing lysimeters. Irrigation science, 28(1), 35-50.
Di Tomaso, J.M. 1998. Impact, biology, and ecology of saltcedar (Tamarix spp.) in the southwestern United States. Weed technology, 326-336.
Figuerola, P.I., & Berliner, P.R. 2005. Evapotranspiration under advective conditions. International journal of biometeorology, 49(6), 403-416.
Fuchs, M., & Tanner, C.B. 1970. Error analysis of Bowen ratios measured by differential psychrometry. Agricultural Meteorology, 7, 329-334.
Gavilán, P., & Berengena, J. 2007. Accuracy of the Bowen ratio-energy balance method for measuring latent heat flux in a semiarid advective environment. Irrigation Science, 25(2), 127-140.
Glenn, E.P., Nagler, P.L., Morino, K., & Hultine, K.R. 2013. Phreatophytes under stress: transpiration and stomatal conductance of saltcedar (Tamarix spp.) in a high-salinity environment. Plant and Soil, 1-18.
Guo, Y., & Schuepp, P.H. 1994. On surface energy balance over the northern wetlands: 2. The variability of the Bowen ratio. Journal of Geophysical Research: Atmospheres (1984–2012), 99(D1), 1613-1621.
Heilman, J.L., Brittin, C.L., & Neale, C.M.U. 1989. Fetch requirements for Bowen ratio measurements of latent and sensible heat fluxes. Agricultural and Forest Meteorology, 44(3), 261-273.
Leclerc, M.Y., & Thurtell, G.W. 1990. Footprint prediction of scalar fluxes using a Markovian analysis. Boundary-Layer Meteorology, 52(3), 247-258.
Lee, X., Yu, Q., Sun, X., Liu, J., Min, Q., Liu, Y., & Zhang, X. 2004. Micrometeorological fluxes under the influence of regional and local advection: a revisit. Agricultural and forest meteorology, 122(1), 111-124.
Mahrt, L., & Ek, M. 1984. The influence of atmospheric stability on potential evaporation. J. Clim. Appl. Meteorol., 23, 222–234.
Monteith, J.L. 1965. Evaporation and environment. In Symp. Soc. Exp. Biol (Vol. 19, No. 205-23, p. 4).
71
Motha, R.P., Verma, S.B., & Rosenberg, N.J. 1979. Exchange coefficients under sensible heat advection determined by eddy correlation. Agricultural meteorology, 20(4), 273-280.
Nagler, P.L., Morino, K., Murray, R.S., Osterberg, J., & Glenn, E.P. 2009. An empirical algorithm for estimating agricultural and riparian evapotranspiration using MODIS Enhanced Vegetation Index and ground measurements of ET. I. Description of method. Remote Sensing, 1(4), 1273-1297.
Papaioannou, G., Jacovides, C., & Kerkides, P. 1989. Atmospheric stability effects on eddy transfer coefficients and on Penman's evaporation estimates. Water resources management, 3(4), 315-322.
Payero, J.O., Neale, C.M.U., Wright, J.L., & Allen, R.G. 2003. Guidelines for validating Bowen ratio data. Transactions of the ASAE, 46(4), 1051-1060.
Perez, P.J., Castellvi, F., Ibanez, M., & Rosell, J.I. 1999. Assessment of reliability of Bowen ratio method for partitioning fluxes. Agricultural and Forest Meteorology, 97(3), 141-150.
Shafroth, P.B., Cleverly, J.R., Dudley, T.L., Taylor, J.P., Riper III, C.V., Weeks, E.P., & Stuart, J.N. 2005. Control of Tamarix in the western United States: implications for water salvage, wildlife use, and riparian restoration. Environmental Management, 35(3), 231-246.
Tanner, B.D., Greene, J.P., & Bingham, G.E. 1987. A Bowen ratio design for long term measurements. American Society of Agricultural Engineers (Microfiche collection)(USA).
Todd, R.W., Evett, S.R., & Howell, T.A. 2000. The Bowen ratio-energy balance method for estimating latent heat flux of irrigated alfalfa evaluated in a semi-arid, advective environment. Agricultural and Forest Meteorology, 103(4), 335-348.
Twine, T.E., Kustas, W.P., Norman, J.M., Cook, D.R., Houser, P., Meyers, T.P., Prueger, J.H., Starks, P.J., & Wesely, M.L. 2000. Correcting eddy-covariance flux underestimates over a grassland. Agricultural and Forest Meteorology, 103(3), 279-300.
USFWS-CNWR. Accessed June 3, 2014. Cibola National Wildlife Refuge | Arizona and California. U.S. Fish & Wildlife Service. <http://www.fws.gov/refuge/Cibola/>
van Gorsel, E., Leuning, R., Cleugh, H.A., Keith, H., Kirschbaum, M.U., & Suni, T. 2008. Application of an alternative method to derive reliable estimates
72
of nighttime respiration from eddy covariance measurements in moderately complex topography. agricultural and forest meteorology, 148(6), 1174-1180.
Verma, S.B., Rosenberg, N.J., & Blad, B.L. 1977. Turbulent exchange coefficients for sensible heat and water vapor under advective conditions. Journal of Applied Meteorology, 17(3), 330-338.
Wilson, K.B., Hanson, P.J., Mulholland, P.J., Baldocchi, D.D., & Wullschleger, S.D. 2001. A comparison of methods for determining forest evapotranspiration and its components: sap-flow, soil water budget, eddy covariance and catchment water balance. Agricultural and Forest Meteorology, 106(2), 153-168.
73
CHAPTER 4
SALTCEDAR EVAPOTRANSPIRATION PRE AND POST FIRE
4.1 Abstract
Bowen ratio and eddy covariance energy balance flux towers have been
installed since 2006 in a dense salt cedar forest within the Cibola National
Wildlife Refuge, California, along the floodplain of the lower Colorado River to
monitor the evapotranspiration of saltcedar (Tamarix ramosissima). Past
research has analyzed data collected from these systems along with information
from groundwater wells, scintillometer readings, and remote sensing approaches.
In the summer of 2011, a wildfire erupted destroying both the saltcedar and
research equipment. To monitor the regrowth after the fire, groundwater
piezometers and two eddy covariance flux towers were reestablished. Additional
measurements of salt cedar water consumption were gathered, augmenting pre-
fire records. Analyzed tower and well measurements revealed a reduction of
saltcedar evapotranspiration and an increase in groundwater elevation post fire.
4.2 Introduction
Water use of saltcedar (Tamarix sp.), specifically along riparian corridors
in the southwestern United States, remains a question despite a half century of
interorganizational studies and collaborations. An exotic but naturalized
facultative phreatophyte, saltcedar has thrived where soil alkalinity, groundwater,
and surface disturbances have coexisted, which has culminated into annual
increases of areal cover and draw on water resources (Brotherson and Winkel,
74
1986; Di Tomaso, 1998). By means of superior competitive abilities, saltcedar
has threatened native plant communities and has been the target of numerous
governmental programs either to slow the advance or eradicate the species.
Many researchers have disagreed as to the predominant means of propagation
(Robinson, 1965; DeLoach et al., 2000; Glenn and Nagler, 2005; Stromberg et
al., 2009) and to the consumptive use of water (summarized in Shafroth et al.,
2005) as compared to the displaced native phreatophytes, both factors having
political and environmental implications. Regardless of the cause, saltcedar now
inhabits every major watershed in the American Southwest (Zouhar, 2003) with
an ever-expanding influence on groundwater resources.
Fire prevalence historically in the Southwest had been low in floodplains
due to presence of water and quick decomposition of potential fuel, but it has
more recently become a greater factor. Evidence points toward saltcedar’s
dense canopy, salt excretions, and resultant leaf litter that increase the possibility
of a fire. Saltcedar stands typically experience burns every 15-20 years, which
when combined with the ability to quickly resprout from the root crown have
created monospecific communities. These characteristics have enabled long-
term survivability and dominance over other native riparian vegetation. (Busch,
1995; Ellis, 2001; Lewis et al., 2003; Zouhar, 2003) Fire frequency impacts both
the surface vegetation and subsurface water supplies by affecting transpiration.
Interestingly, a study concluded that saltcedar exhibited higher midday
transpiration rates in recovering burned specimen than in controls (Busch and
Smith, 1993). If this were the case (as this subject is the foundation of this
75
paper), then to what extent, if any, would a burn impact saltcedar water use? In
answer to this question, the following analysis determines to compare saltcedar
evapotranspiration before and after a burn in the United States Southwest.
Groundwater and evapotranspiration research data of a saltcedar-infested
region bordering the Colorado River in southern California has been monitored
by groundwater and flux tower measurements since 2006. A wildfire that
ravaged the area in 2011 became a unique opportunity to conduct research,
because it transformed an actively transpiring, mature saltcedar forest within a
short interval into a blackened and barren landscape. This afforded an instant
comparison to measure the impact of no vegetation on the groundwater.
Additionally, the effects of the resurgent growth on groundwater levels were
examined. Altogether, the burn enabled a singular verification of measured
fluxes and modeled energy and water balances at the measurement sites.
4.3 Methods
Plant groundwater consumption may be estimated by measuring the
change in the phreatic surface elevation. By accounting for mass source fluxes
of inflow and outflow, the remaining dimension equates to the volumetric uptake
within the medium porosity. To determine the actual volume, the ratio of the
affected void to the total volume is multiplied. This last soil property, the specific
yield, must be known or estimated. A mathematical representation for
groundwater evapotranspiration (ET) over a period of time is presented in
Equation 4.1 below.
76
)( HRSET YG ∆−= [ 4.1 ]
where GET is groundwater evapotranspiration, YS is specific yield, R is
the change in depth due to lateral recharge or depletion, and H∆ is the
change in phreatic surface elevation.
By measuring groundwater fluctuations in a well, one of the two inner
variables in Equation 4.1 is readily obtained, which then leaves the recharge.
White (1932) documented a method, which has been widely applied, to estimate
the recharge rate for a day as the mean groundwater fluctuation for the first few
hours after midnight assuming no plant transpiration during this period. Since the
estimated day-to-day recharge rates vary sporadically and the assumption of no
nighttime transpiration contested, an alternate method was considered. Soylu et
al. (2012) proposed to calculate the recharge by detrending the oscillating diurnal
groundwater movements. Because it forms a sinusoidal function, the method
was called a “Fourier-based improvement” to the general White Method and is
replicated in Equation 4.2. C or the amplitude is functionally the recharge minus
the change in elevation deduced by removing the offset, linear, and wave trends.
This equation may be applied to time steps of periods involving one or more
days.
))(2sin()( Dtt
CBtAtFN
+++=π [ 4.2 ]
77
where )(tF is the groundwater stage at a given time t , A is the scalar
offset, B is the linear offset, C is the amplitude, Nt is the number of
increments per cycle or day, and D is the phase offset.
By derivation, C is one-half of HR ∆− , which is subsequently doubled and
multiplied by a corrective factor (an artifact of the detrending process) to estimate
groundwater ET as shown in Equation 4.3. Still required is the specific yield, but
Equation 4.3 takes on a different form than Equation 4.1 where the two
corrections may be treated as one variable and obtained through calibration.
Other transpiration measurement techniques, such as lysimetry, eddy
covariance, Bowen ratio, chambers, or sap flow, could be used for this purpose.
))(2( CkSET YG = [ 4.3 ]
where k is a scaling factor resulting from the detrending operations.
4.4 Materials
Situated along the Lower Colorado River in the middle of the Sonoran
Desert, the Cibola National Wildlife Refuge (CNWR) is preserved by the U.S.
Fish and Wildlife Service as a haven for migratory birds and other wildlife
(USFWS-CNWR 2014). CNWR encompasses the river floodplain of California
on the west and Arizona on the east and has been overrun by saltcedar (Tamarix
ramosissima). Since 2006, several universities in connection with the U.S.
Bureau of Reclamation have been investigating ET rates within the saltcedar
forests by recording sap flux, groundwater fluctuations, and tower measurements
78
using scintillometer, eddy covariance (EC), Bowen ratio (BR), and Penman
techniques. Early measurements were analyzed and published, including
groundwater ET via the standard White Method (Taghvaeian 2011). However,
an unforeseen but ultimate benefit to the study occurred when the Three Slashes
Fire burned the saltcedar in late August 2011. Because most of the equipment
had expired in the inferno, towers were reconstructed and well caps replaced in
early 2012. Measurements resumed, and the data collected characterized the
water use of the reemerging saltcedar. Although instruments and hence capital
were lost in the flames, the overarching goal of measuring saltcedar impact on
the Colorado River-supplied groundwater was enhanced by being able to
evaluate the effects of the absence of saltcedar in the floodplain. Such a
scenario would never have been permitted in the study scope, but from the point
of view of this study the fire provided an opportunity for a “natural experiment”.
Five false-color aerial photos (NAIP 2014) including the three tower sites
in the study area are shown in Fig. 4.1, which depict the saltcedar growth well
before and soon after the fire and also the regrowth. Nagler et al. (2009b)
described the study area pre fire and indicated that the plain was dominated by
invasive saltcedar with occasional patches of native arrowweed (Pluchia
sericea), mesquite (Prosopis pubescens), and quailbush (Atriplex lentiformis)
intermingled. Correspondingly, inspection post fire included the same vegetation
patterns with the exception of arrowweed covering some wide stretches. Of note
in Fig. 4.1.D is a region where the fire did not advance that had already
experienced a fire in 2006, and careful inspection of subfigure B shows the
79
regrowth. Analysis of the 2006 blaze found that a previous fire had burned
CNWR in 1991 with the sample site nearest the river experiencing the greatest
intensity (Godaire and Klinger, 2007). Confirmed at both locations precisely was
the 15-20 year proposed fire interval for saltcedar.
Fig. 4.1. False color (infrared, red, green) images from the National Agriculture Imagery Program (NAIP) of the tower sites in the Cibola National Wildlife Refuge, (A) June 26, 2005; (B) May 25, 2009; (C) April 26 and 28, 2010; (D) April 23, 2012; (E) May 16, 2014.
80
Tower sites were positioned in a dense stand of salt cedar (Slithern),
medium-density salt cedar (Diablo), and next to the river (Swamp), which were at
least the circumstances pre fire. Groundwater monitoring stations were located
next to each tower site with an additional four wells in the near proximity (Fig.
4.2). Once elevation differences were eliminated among well sites through a
high-resolution LIDAR-derived surface elevation model and considering the
overall accuracy of the groundwater loggers, measurements from the clusters of
five well sites were nearly identical (Taghvaeian, 2011). This redundancy proved
vital as the groundwater data acquisition was able to survive fire, battery failures,
and change in technicians with few gaps. Beyond the potential for fire, CNWR
has a harsh climate with summer temperatures exceeding 50 °C and desert
conditions of 80 mm average precipitation annually (of which most falls in the
winter months). Because of the extreme environment, maintenance once every
1-2 months due to site remoteness, and non-redundant tower instruments, data
gaps did occur in the surface flux measurements but were remedied as quickly
as possible.
As Fig. 4.1 and Fig. 4.2 both show, the stream nearest the station sites is
actually the old channel of the Colorado River with the engineered channel on
the east, and it is fed both by the Palo Verde Irrigation District outlet drain and
the river further upstream. The primary source of groundwater in the study area
of CNWR is the drain, and higher groundwater salinity has been observed at the
wells further from the source. Regrowth post fire around Slithern tower site is
shown in Fig. 4.3 with a high percentage of saltcedar reemerging and achieving
81
significant growth in just two years. Although not specifically sampled, the
resprouting from the root crown exceeded the 53 and 55% reported by Ellis
(2001) and echoed the observation of Zouhar (2003) that saltcedar regrowth “can
reach several feet in just a few months”.
Fig. 4.2. Tower and groundwater well observation locations within the Cibola National Wildlife Refuge.
82
Fig. 4.3. Saltcedar regrowth in the Cibola National Wildlife Refuge post fire (From Top to bottom, March 12, 2012; May 29, 2013; August 20, 2013).
4.5 Data Analysis and Results
BR and EC tower measurements were examined and corrected with the
methodology outlined in a previous publication (Lewis and Neale, 2014). It was
found that horizontal heat advection from the surrounding desert seriously
83
impaired BR ET before and after the fire. Attempts at compensating for
advection through EC-based corrections for the BR data are shown in Fig. 4.4
(on a weekly moving daily average to increase clarity of the trends). Since
Swamp never had an EC setup, its BR data and resultant ET were not corrected
and are therefore not included. What can be gleaned from the BR ET are
seasonal patterns with higher water use shifted at Slithern to later in the year
after the fire. It additionally highlights differences between an energy balance
method and a mass balance (in a later figure). Presentation of the BR results
also is important because of the many papers that have used the data previously
for comparison to sap flow, remote sensing, groundwater, and scintillometers
(Nagler et al., 2009a; Taghvaeian, 2011; Geli et al., 2012). This was mainly
because the method was common to all towers, and because the EC data had
never been fully processed. Shown in Fig. 4.5 is the weekly averaged daily EC
ET for Diablo pre fire (which has a substantial record) and for Diablo and Slithern
post fire. Energy balance closure of the EC data for Diablo and Slithern was
around 70 percent (not untypical for the environment), and the remaining energy
was proportioned to latent and sensible heat fluxes by using the Bowen ratio
(Twine et al., 2000). ASCE standardized short reference ET (Walter et al., 2000)
in Fig. 4.4, Fig. 4.5, and later figures was calculated from hourly data at a local
electronic weather station (Palo Verde II; CIMIS 2014) with missing
meteorological data filled in with adjusted values from a NASA gridded weather
forcing model (NLDAS-2).
84
Fig. 4.4. Weekly averaged daily ET from Bowen ratio systems in the Cibola National Wildlife Refuge, (Top: 2006-2010; Bottom: 2012-2013).
Preparation of the groundwater pressure and temperature readings
involved initial data screening, subtraction of atmospheric pressure, conversion
to depths, and determination of phreatic surface elevation (above mean sea
level) through well cord lengths and the LIDAR dataset that established the exact
ground elevation at each well. Readings were logged at 15-minute intervals and
were slightly offset at different well sites. These values were shifted to regular
0
2
4
6
8
10
1 31 61 91 121 151 181 211 241 271 301 331 361
0
2
4
6
8
10
1 31 61 91 121 151 181 211 241 271 301 331 361
Diablo Slithern Swamp Reference ET
Eva
potra
nspi
ratio
n (m
m/d
ay)
85
increments or to :00, :15, :30, and :45 minutes of each hour through linear
interpolation. Afterwards, pressures and temperatures were averaged for each
hour, which was the time step used for subsequent calculations. Atmospheric
pressure over CNWR was obtained by adjusting the continuous record of the
Blythe (California) airport weather station to match barometric readings from two
suspended, unsubmerged loggers in wells at Slithern tower and a nearby well
location. Because adjacent readings were analogous, groundwater elevations
were grouped by tower site and averaged. In the process, many errors could be
detected through comparison (e.g., abrupt changes in cord length after a logger
had been retrieved and replaced, bad data, incorrectly entered offsets) and either
be corrected or excluded from the averaging. Further, loss of data through
burned out wells, dead batteries, and human errors was mitigated by the similar
readings at adjacent well sites.
Final averaged groundwater elevations for each tower group are plotted in
Fig. 4.6 likewise on a weekly moving daily average. Stream elevation shown
was provided by a U.S. Bureau of Reclamation gaging station, Palo Verde Outfall
Drain, which is roughly 9 kilometers upstream from the point on the drain nearest
Slithern. Seasonal patterns of phreatophyte ET are evident as the groundwater
table recovers in the winter from summertime lower levels. Diablo and Slithern
groundwater elevations are comparable with lower elevations at Diablo expected
by its further distance inland. Swamp, being closer to the source, is affected by
the day-to-day fluctuations of the water levels in the drain but also does not drop
as low during the peak ET season. Post-fire groundwater recharge attained
86
unprecedented levels in the record and lowered only a fraction of what pre-fire
levels were during the growing season. With stream stage or inflow maintaining
typical behavior, an initial conclusion would be that surface vegetative
transpiration requirements or groundwater ET had lessened post fire.
Fig. 4.5. Weekly averaged daily ET from eddy covariance systems in the Cibola National Wildlife Refuge, (Top: 2007-2009; Bottom: 2012-2013).
0
2
4
6
8
10
1 31 61 91 121 151 181 211 241 271 301 331 361
0
2
4
6
8
10
1 31 61 91 121 151 181 211 241 271 301 331 361
Diablo Slithern Reference ET
Eva
potra
nspi
ratio
n (m
m/d
ay)
87
Amplitudinal coefficients were calculated from a least-squares fit of the
group averaged groundwater elevations to Equation 4.2. Because of the nature
of the detrending, variable lengths of data could be applied to the equation,
ranging normally from a day to a week or more. A moving calculation period of
25 hours was selected with 12 hours before and after the record of interest
(example fits in Fig. 4.7). Output was technically hourly but represented a much
longer duration. Final amplitudes were averaged for each day and doubled in
preparation to be multiplied by additional factors of specific yield and the
detrending correction in Equation 4.3.
Groundwater ET was calculated using the Fourier-based White Method
(FW) near each tower site, and the product of the specific yield and scaling factor
was determined from a least-squares fit to overlapping latent heat flux from the
EC method at Diablo and Slithern. Both EC and BR are frequently unreliable at
night where saltcedar has been documented to transpire at 12-25% or more of
total daily ET (Dawson et al., 2007; Nagler et al., 2009b; Hultine et al., 2010;
Glenn et al., 2013). Since the EC calculations displayed no ET at night, daily
estimated ET was increased by a value of one-third. This addition corresponds
to reported nighttime transpiration totaling 25% of daily values from sap flux
sensors in a previous study (Nagler et al., 2009b). Calibration for Swamp, which
never had an EC system, was obtained by matching the wintertime baseline
groundwater ET of Diablo and Slithern pre fire. Pseudo (by incorporation of the
corrective factor) specific yield values were between 0.107 and 0.125 for all three
sites (Table 4.1, which also summarizes other site properties). Well site
88
Fig. 4.6. Weekly average daily groundwater surface elevation from multiple wells at 3 sites in the Cibola National Wildlife Refuge.
Fig. 4.7. Example of 24 hour Fourier-White method implementation for 3 days at Diablo tower well, Cibola National Wildlife Refuge (July 1-7, 2011).
63.0
63.5
64.0
64.5
65.0
65.5
66.0
66.5
2006 2007 2008 2009 2010 2011 2012 2013
Ele
vatio
n (m
eter
)
Diablo Slithern Swamp Stream Stage Fire
63.40
63.45
63.50
63.55
63.60
7/1/2011 7/2/2011 7/3/2011 7/4/2011 7/5/2011 7/6/2011 7/7/2011
Ele
vatio
n (m
eter
)
Diablo F(t)
89
characterizations were obtained pre fire and included water quality testing and
soil samples. At a depth of 2-4 meters—which is where the groundwater table
fluctuated—and for all sampled soil cores near Diablo and Slithern, the texture
classification was either a sandy-loam or a loamy-sand. According to Johnson
(1967), specific yields for a corresponding grain size of fine sand are between
0.10 and 0.28 with 0.21 being the average. Loheide et al. (2005) defined a
‘readily available specific yield’ as “a function of drainage time, depth to the water
table, and the properties of the porous media” for which fluctuations of the water
table do not occur instantaneously as is suggested in a laboratory setting. These
in-field values of specific yield are therefore a fraction of the expected value. In
his work, Loheide and his associates found that for large grain sizes the readily
available specific yield was nearly equal the specific yield but for smaller grain
sizes the fraction could be as low as 40%. Assuming the averaged value for
specific yield in the study site of 0.21 and Slithern’s calculated (pseudo) specific
yield of 0.125 as the readily available specific yield, the fraction of the two would
yield 60%. This is low for the site’s sand and loam texture but not unreasonable
as it can be presupposed by a lower-than-average specific yield or a k value less
than one.
Weekly averaged daily groundwater ET for the three well groups together
with calculated short reference ET is plotted in Fig. 4.8 with years before the fire
and after the fire averaged for simplicity. Both Diablo and Slithern wells exhibited
very good agreement in the pre-fire period while groundwater ET at Swamp was
strikingly higher and followed reference ET levels during the non-dormant period.
90
Fig. 4.8. Weekly averaged daily groundwater et from multiple wells at 3 sites in the Cibola National Wildlife Refuge, (From Top to Bottom, 2006-2010; 2011; 2012-2013).
Gro
undw
ater
Eva
potra
nspi
ratio
n (m
m/d
ay)
91
Table 4.1. Averaged groundwater properties for multiple wells at 3 sites in the Cibola National Wildlife Refuge pre and post fire.
Parameter Period Unit Diablo Slithern Swamp Specific Yield 2006-2013 — 0.113a 0.125a 0.107b
Depth from Surface 2006-2013 meter 2.96 3.33 2.62 Temperature 2006-2013 °C 22.4 21.8 22.2
Electrical Conductivity 2006-2011 mS/cm 16.58 8.30 3.95 Note: Specific yield aderived from calibration to eddy covariance measurements and bestimated by matching wintertime baseline to Diablo and Slithern.
Previously, suspicion was directed towards the validity of Swamp groundwater
ET as it heavily correlated to the rise and fall of the drain stage. Yet, the fire in
August 2011 confirmed the absence of any direct coupling of the drain flow (in
addition to feeding the groundwater) on the calculated groundwater ET for all
three well groups by the sudden and sustained drop in groundwater ET post fire,
which abruptness is moderated by the weekly average in the figure.
Post fire results are congruent with previous states, and the effect of the
burn was reduced groundwater ET rates. Only at Slithern was an annual offset
present which otherwise mimicked Diablo groundwater ET. One possible
explanation is that Slithern wells were not measuring groundwater consumption
of sole saltcedar regrowth. In fact, sizeable swaths of native arrowweed, an
evergreen (Ferren et al., 1996), had occupied some areas not too far eastward of
Slithern tower. This could also explain the non-zero usage during dormant
saltcedar periods in previous years. Averaged daily groundwater ET summed to
92
annual values, both up to the fire and after, is presented in Table 4.2. Both
Diablo and Swamp have lower rates post fire at different scales, but Slithern
groundwater ET actually increases after the fire. This could be the outcome of a
greater presence of arrowweed among the saltcedar.
4.6 Discussion
Both pre and post fire, results from this study indicate that the saltcedar-
groundwater relationship in CNWR was affected by surface-to-groundwater
depth, salinity, plant phenology, and the fire. Moreover, our observations were in
agreement with Devitt et al. (1998) that “[saltcedar] has the potential to be both a
low water user and a high water user, depending on water availability, canopy
development, and atmospheric demand, and that advection can dominate energy
balances and ET in arid land riparian zones”. What was not observed were soil
hydraulic limitations (Glenn et al., 2013) and maximum or potential rates of
saltcedar ET. Overall, groundwater depth and salinity at Diablo, Slithern, and
Swamp were more important factors than surface vegetative conditions pre fire,
while post fire groundwater was not the limiting factor.
Table 4.2. Average annual groundwater ET from multiple wells at 3 sites in the Cibola National Wildlife Refuge Pre and post fire assuming conservative estimates of specific yield (meter/year).
Period Diablo Slithern Swamp Ref. ET Pre Fire (2006-2011) 0.805 0.838 1.923 1.786 Post Fire (2011-2013) 0.327 0.909 1.331 1.773
93
Previous studies in CNWR (Taghvaeian, 2011; Glenn et al., 2013)
concluded that depth to groundwater constrained saltcedar ET but could not
explain the dissimilarities among BR results and yet comparable groundwater ET
at Diablo and Slithern. Further exacerbation stemmed from Swamp groundwater
readings that revealed more than double annual groundwater ET rates at a site
only a few hundred meters closer to the drain than Slithern. Yet, the average
depths to groundwater were lowest at Swamp and 0.34 and 0.71 m higher at
Diablo and Slithern, respectively. This is in harmony with the difference seen in
saltcedar ET by Van Hylckama (1970) in groundwater depth changes of 1.5
meters to 2.1 meters and 2.1 meters to 2.7 meters. Reported shifts in saltcedar
water uptake from the phreatic to the vadose zone (Nippert et al., 2010) were not
apparent, for when the groundwater maintained a constant depth so did the
groundwater ET. This is most likely due to the lack of precipitation during the
midsummer months or a buildup of salts in the vadose zone as proposed by
Glenn et al. (2013).
Practically equal estimates of groundwater ET from measurements at
Diablo and Slithern pre-fire and longer depth to groundwater at Slithern would
void the argument for higher groundwater ET rates at Swamp if based solely on
depth to groundwater. Averaged measured groundwater electrical conductivity
was half the value at Slithern than at Diablo and less than half again at Swamp.
Both saltcedar and arrowweed have been found to tolerate the groundwater
salinities reported at the three locations in this study, but reduction of
transpiration due to higher salinity was also reported with more gradual rates for
94
saltcedar than for arrowweed (Glenn et al., 1998; Vandersande et al. 2001).
Given the opposite changes in depth to groundwater and salinity at both Diablo
and Slithern, similar groundwater ET rates should not be surprising.
Assuming the maximum specific yield of 0.28 for the sandy and loamy soil
in CNWR and that the reported annual groundwater ET for each location was
based on a specific yield of 0.10, the factor controlling the maximum probable
estimates would be 0.28 / 0.10 = 2.8. By applying this upper bound, annual
groundwater ET pre fire would range between 1.87 and 5.05 meters/year, the
original estimates being 0.81 and 1.92 meters/year. All but the highest estimate
fall within the ET rates suggested in the literature (reviewed by Shafroth et al.,
2005) and which confirms the reduction of the readily available specific yield.
More recent studies (Dahm et al., 2002; Nagler et. al., 2008; Westenburg et al.,
2006) support the eddy covariance-matched rates in Table 4.2 while the results
of Sala et al. (1996) of saltcedar ET up to twice the potential rate would
correspond with maximum specific yields. Based on evidence from the
meteorological tower measurements, this study backs the latest figures.
Reduction in saltcedar ET is a direct relationship to how well the
environment duplicates theoretical or greenhouse conditions. Groundwater ET at
Swamp visibly follows the rise and fall of the drain stage influenced by the
Colorado River and would transpire at greater rates if the drain always
maintained a higher level (in line with Devitt et al., 1997a and 1997b). The same
is true at Diablo and Slithern pre fire with late spring surges in groundwater ET
reversed by groundwater depressions. This phenomenon may also be
95
influenced by the saltcedar growth patterns, which show higher consumption in
the earlier portion of the year with a downward trend throughout the rest of the
season as evidenced before and after the fire. Following the 2011 fire,
groundwater ET became nearly zero with likely some plants that never burned
still transpiring. Fig. 4.1.D and E clearly show that post fire regrowth was much
higher around Slithern than at Diablo and corroborate the higher annual
groundwater ET post fire at Slithern while still maintaining higher groundwater
levels from lower groundwater ET rates elsewhere. Part of these annual values
was already allocated to much greater arrowweed contribution. If the offset of
1.6 mm/day were subtracted, then the annual total would become 0.324
meters/year or parallel to Diablo. This would apportion saltcedar groundwater ET
at Diablo and Slithern at 40 percent of pre-fire rates and Swamp at 70 percent.
4.7 Conclusions
Due to a 2011 burn in the CNWR, measured saltcedar water use dropped
at three locations near an agricultural drain feeding into the lower Colorado River.
Groundwater recovery was immediately evident and sustained for the next two
years. Study results concur with Di Tomaso (1998) and DeLoach et al. (2000)
that removal of saltcedar yields increases in water availability. At least in the
case of CNWR, frequent burns of the dense saltcedar forest would most likely
decrease inter-annual ET, thus providing more water downstream but would be
unlikely to eliminate the infestation. Magnitudes of saltcedar water use were
within the realm of native phreatophytes. Since measurements are ongoing,
96
future data analysis of the effects of the Mediterranean tamarisk beetle
(Diorhabda elongata, Lewis et al. 2003) on saltcedar groundwater consumption
will be interesting as the beetle migrates south along the Colorado River corridor
and will soon be present at CNWR.
4.8 Acknowledgments
This study was supported by the Six Agency Committee of the Colorado
River and the U.S. Bureau of Reclamation through a cooperative agreement.
Partial support was also provided by the Utah Agricultural Experiment Station
and the Remote Sensing Services Laboratory in the Department of Civil and
Environmental Engineering at USU.
4.9 References
Brotherson, J.D., & Winkel, V. 1986. Habitat relationships of saltcedar (Tamarix ramosissima) in central Utah. Western North American Naturalist, 46(3), 535-541.
Busch, D.E. 1995. Effects of fire on southwestern riparian plant community structure. The Southwestern Naturalist, 259-267.
Busch, D.E., & Smith, S.D. 1993. Effects of fire on water and salinity relations of riparian woody taxa. Oecologia, 94(2), 186-194.
CIMIS. Accessed August 28, 2014. California Irrigation Management Information System | California Department of Water Resources, University of California at Davis. <http://www.cimis.water.ca.gov/cimis/welcome.jsp>.
Dahm, C.N., Cleverly, J.R., Allred Coonrod, J.E., Thibault, J.R., McDonnell, D.E., & Gilroy, D.J. (2002). Evapotranspiration at the land/water interface in a semi‐arid drainage basin. Freshwater Biology, 47(4), 831-843.
Dawson, T.E., Burgess, S.S., Tu, K.P., Oliveira, R.S., Santiago, L.S., Fisher, J.B., Simonin, K.A., & Ambrose, A.R. 2007. Nighttime transpiration in woody plants from contrasting ecosystems. Tree Physiology, 27(4), 561-575.
97
DeLoach, C.J., Carruthers, R.I., Lovich, J.E., Dudley, T.L., & Smith, S.D. 2000, July. Ecological interactions in the biological control of saltcedar (Tamarix spp.) in the United States: toward a new understanding. In N. R. Spencer (Ed.), Proceedings of the X international symposium on biological control of weeds (Vol. 20001, pp. 819-8731). Bozeman^ eMontana Montana: Montana State University.
Devitt, D.A., Piorkowski, J.M., Smith, S.D., Cleverly, J.R., & Sala, A. 1997a. Plant water relations of Tamarix ramosissima in response to the imposition and alleviation of soil moisture stress. Journal of Arid Environments, 36(3), 527-540.
Devitt, D.A., Sala, A., Mace, K.A., & Smith, S.D. 1997b. The effect of applied water on the water use of saltcedar in a desert riparian environment. Journal of Hydrology, 192(1), 233-246.
Devitt, D.A., Sala, A., Smith, S.D., Cleverly, J., Shaulis, L.K., & Hammett, R. 1998. Bowen ratio estimates of evapotranspiration for Tamarix ramosissima stands on the Virgin River in southern Nevada. Water Resources Research, 34(9), 2407-2414.
Di Tomaso, J.M. 1998. Impact, biology, and ecology of saltcedar (Tamarix spp.) in the southwestern United States. Weed technology, 326-336.
Ellis, L.M. 2001. Short-term response of woody plants to fire in a Rio Grande riparian forest, Central New Mexico, USA. Biological Conservation, 97(2), 159-170.
Ferren, W.R., Fiedler, P.L., Leidy, R.A., Lafferty, K.D., & Mertes, L.A. 1996. Wetlands of California: part 2. Classification and description of wetlands of the central and southern California coast and coastal watersheds. Madroño, 43(1), 125-182.
Geli, H.M., Neale, C.M., Watts, D., Osterberg, J., De Bruin, H.A., Kohsiek, W., Pack, R.T., & Hipps, L.E. 2012. Scintillometer-Based Estimates of Sensible Heat Flux Using Lidar-Derived Surface Roughness. Journal of Hydrometeorology, 13(4), 1317-1331.
Glenn, E.P., & Nagler, P.L. 2005. Comparative ecophysiology of Tamarix ramosissima and native trees in western US riparian zones. Journal of Arid Environments, 61(3), 419-446.
Glenn, E.P., Nagler, P.L., Morino, K., & Hultine, K.R. 2013. Phreatophytes under stress: Transpiration and stomatal conductance of saltcedar (Tamarix spp.) in a high-salinity environment. Plant and soil, 371(1-2), 655-672.
98
Glenn, E., Tanner, R., Mendez, S., Kehret, T., Moore, D., Garcia, J., & Valdes, C. 1998. Growth rates, salt tolerance and water use characteristics of native and invasive riparian plants from the delta of the Colorado River, Mexico. Journal of Arid Environments, 40(3), 281-294.
Godaire, J.E., Klinger, R.E. 2007. Tree-Ring Dating of Tamarisk (Tamarix ramosissima), Cibola National Wildlife Refuge, California. U.S. Department of the Interior. Bureau of Reclamation.
Hultine, K.R., Nagler, P.L., Morino, K., Bush, S.E., Burtch, K.G., Dennison, P.E., Glenn, E.P., & Ehleringer, J.R. 2010. Sap flux-scaled transpiration by tamarisk (Tamarix spp.) before, during and after episodic defoliation by the saltcedar leaf beetle (Diorhabda carinulata). Agricultural and Forest Meteorology, 150(11), 1467-1475.
Johnson, A.I. 1967. Specific yield: compilation of specific yields for various materials. US Government Printing Office.
Lewis, C.S. and Neale, C.M. 2014. Bowen Ratio System Assessment over Saltcedar under Different Atmospheric Stability and Advective Conditions by the Eddy Covariance Technique. Unpublished manuscript.
Lewis, P.A., DeLoach, C.J., Knutson, A.E., Tracy, J.L., & Robbins, T.O. 2003. Biology of Diorhabda elongata deserticola (Coleoptera: Chrysomelidae), an Asian leaf beetle for biological control of saltcedars (Tamarix spp.) in the United States. Biological control, 27(2), 101-116.
Loheide, S.P., Butler, J.J., & Gorelick, S.M. 2005. Estimation of groundwater consumption by phreatophytes using diurnal water table fluctuations: a saturated‐unsaturated flow assessment. Water resources research, 41(7).
Nagler, P.L., Glenn, E.P., Didan, K., Osterberg, J., Jordan, F., & Cunningham, J. 2008. Wide‐Area Estimates of Stand Structure and Water Use of Tamarix spp. on the Lower Colorado River: Implications for Restoration and Water Management Projects. Restoration Ecology, 16(1), 136-145.
Nagler, P.L., Morino, K., Murray, R.S., Osterberg, J., & Glenn, E.P. 2009b. An empirical algorithm for estimating agricultural and riparian evapotranspiration using MODIS Enhanced Vegetation Index and ground measurements of ET. I. Description of method. Remote Sensing, 1(4), 1273-1297.
NAIP. Accessed June 3, 2014. National Agriculture Imagery Program | Cal-Atlas Geospatial Clearinghouse. <https://projects.atlas.ca.gov/projects/naip/>
99
Nippert, J.B., Butler Jr, J.J., Kluitenberg, G.J., Whittemore, D.O., Arnold, D., Spal, S.E., & Ward, J.K. 2010. Patterns of Tamarix water use during a record drought. Oecologia, 162(2), 283-292.
NLDAS-2. The North American Land Data Assimilation System Phase 2. NASA Goddard Space Flight Center and Other Collaborators. <ftp://hydro1.sci.gsfc.nasa.gov/data/s4pa//NLDAS/NLDAS_FORA0125_H.002/>.
Robinson, T.W. 1965. Introduction, spread and areal extent of saltcedar (Tamarix) in the western states. US Government Printing Office.
Sala, A., Smith, S.D., & Devitt, D.A. 1996. Water use by Tamarix ramosissima and associated phreatophytes in a Mojave Desert floodplain. Ecological Applications, 888-898.
Shafroth, P.B., Cleverly, J.R., Dudley, T.L., Taylor, J.P., Riper III, C.V., Weeks, E.P., & Stuart, J.N. 2005. Control of Tamarix in the western United States: implications for water salvage, wildlife use, and riparian restoration. Environmental Management, 35(3), 231-246.
Soylu, M.E., Lenters, J.D., Istanbulluoglu, E., & Loheide, S.P. 2012. On evapotranspiration and shallow groundwater fluctuations: A Fourier‐based improvement to the White method. Water Resources Research, 48(6).
Stromberg, J.C., Chew, M. K., Nagler, P.L., & Glenn, E.P. 2009. Changing perceptions of change: the role of scientists in Tamarix and river management. Restoration Ecology, 17(2), 177-186.
Taghvaeian, S. 2011. Water and energy balance of a riparian and agricultural ecosystem along the Lower Colorado River (Doctoral dissertation, UTAH STATE UNIVERSITY).
Twine, T.E., Kustas, W.P., Norman, J.M., Cook, D.R., Houser, P., Meyers, T.P., Prueger, J.H., Starks, P.J., & Wesely, M.L. 2000. Correcting eddy-covariance flux underestimates over a grassland. Agricultural and Forest Meteorology, 103(3), 279-300.
USFWS-CNWR. Accessed June 3, 2014. Cibola National Wildlife Refuge | Arizona and California. U.S. Fish & Wildlife Service. <http://www.fws.gov/refuge/Cibola/>
Van Hylckama, T.E. 1970. Water use by salt cedar. Water Resources Research, 6(3), 728-735.
100
Vandersande, M.W., Glenn, E.P., & Walworth, J.L. 2001. Tolerance of five riparian plants from the lower Colorado River to salinity drought and inundation. Journal of Arid Environments, 49(1), 147-159.
Walter, I.A., Allen, R.G., Elliott, R., Mecham, B., Jensen, M.E., Itenfisu, D., Howell, T.A., Snyder, R., Brown, P., Echings, S., Spofford, T., Hattendorf, M., Cuenca, R.H., Wright, J.L., Martin, D., 2000. ASCE’s standardized reference evapotranspiration equation. In: Proc. of the Watershed Management 2000 Conference, June.
Westenburg, C.L., Harper, D.P., & DeMeo, G.A. 2006. Evapotranspiration by Phreatophytes Along the Lower Colorado River at Havasu National Wildlife Refuge, Arizona. US Geological Survey.
White, W.N. 1932. A method of estimating ground-water supplies based on discharge by plants and evaporation from soil: Results of investigations in Escalante Valley, Utah. US Government Printing Office.
Zouhar, K. 2003. Tamarix spp. Fire Effects Information System,[Online]. US Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). Available: http://www.fs.fed.us/database/feis.
101
CHAPTER 5
SUMMARY AND CONCLUSIONS
Mass transfer of water on the earth’s surface through solid, liquid, and
vapor phases has been studied for millennia (Brutsaert, 1982). This has been
partly to facilitate in planning for the efficient construction and service of water
resources projects to provide irrigation, drainage, dams, other control structures,
conveyance, treatment, and potable and industrial use. Design, creation, and
maintenance of these structures and activities are planned for with available raw
materials, engineering expertise, contractors, and workforce. Cost estimates and
design flows can be calculated to some degree of certainty whereas system
removals by evaporation and transpiration remain presumptions backed by
disproportionate data (not to mention input precipitation). Advances in
measurement and prediction of evapotranspiration have occurred in the past
century through development of better evapotranspiration equations,
instrumentation, data handling, and aerial and satellite imagery. While
improvement has been continual, current evapotranspiration estimates are
beneath existing and unimplemented potential and knowhow.
Contributing to this imbalance, no doubt, are several issues. Budget
limitations, bureaucracy, and the lack of vision by its leaders are among the
foremost. As a late example, the Global Precipitation Measurement Mission
priced $978 million (GPM, 2014) and affords global coverage every 2-4 hours.
Landsat missions, equivalent in expense, have much higher resolution but a
102
return of once every 16 days (Landsat, 2014). What would it cost to mass
produce these satellites in the thousands scale to produce images every hour or
better? There would be those who would reply with an astronomical amount.
However, sound reasoning dictates an affordable price if past free market history
can be replicated (e.g., automobiles, computers, appliances, etc.). Results of
such an endeavor would be able to track water transfers on the earth’s surface
with heightened confidence.
Closer to home are legal wars over water and those privy to and against
greater water measurement. Military and privacy concerns might also prohibit
greater collection. Yet, increased knowledge of, which is acquired by
measurement of, evapotranspiration and other water processes is the only way
to more fairly and competently regulate shared water resources. In the end,
engineers and researchers grapple with the inadequate data at hand to stretch
broadened conclusions on the whole. True, current analysis is improved over
yesterday, but it still is lacking in reaching an asymptote on the reliability curve.
With better data, researchers will be able to publish figures instead of
guestimates.
5.1 Objective Fulfillment
NLDAS or the North American Land Data Assimilation System, a
historical dataset of gridded hourly weather parameters centered on the
continental United States that was derived from space, aerial, and land-based
measurements (NLDAS-2), was analyzed in Chapter 2 against an array of
103
electronic weather stations from various state and regional networks. Designed
to monitor and predict drought, NLDAS incorporates all microclimates, from
which many do not represent reference evapotranspiration conditions in an arid
environment. Hence, agricultural networks or those weather stations sited in
representative conditions from other networks were compared against NLDAS for
air temperature, relative humidity, solar radiation, wind speed, and calculated
short reference evapotranspiration. Input parameters and calculated reference
evapotranspiration from NLDAS were verified against the electronic weather
station data to verify model results. Strengthened by well-correlating air
temperature and solar radiation, reference evapotranspiration from modeled
NLDAS matched output from 704 weather stations within an accepted range of
accuracy. Although limited to the western 17 states, NLDAS annual bias was
only slightly higher than the weather station totals for most of the area.
Even though objectification was introduced through interpolations,
microscaled comparisons were not too different from an aggregated model for
hourly reference conditions. Evapotranspirative analysis over large areas is
typically calculated on a daily or instantaneously-extrapolated-to-daily time step.
Findings that NLDAS modestly compliments weather station observations
reinforce its integrity at depicting actual surface conditions. This yields a 24x
resolution over the norm that covers an expansive area in a dataset with no
missing data. From the standpoint of this study, NLDAS may be substituted for
weather station data in many areas of the Intermountain West with acceptable
accuracy.
104
Saltcedar evapotranspiration was studied in Chapters 3 and 4 in the
Cibola National Wildlife Refuge, California. Within the floodplain of the Colorado
River, meteorological instruments were positioned on towers above stands of
saltcedar with groundwater depth loggers monitoring wells below the surface. As
an exotic and proliferating species (genus Tamarix), water use of saltcedar has
been investigated by many researchers to determine comparisons to native
riparian vegetation. Due to a fire in 2011, this study reported transpiration rates
of full growth saltcedar and then of the regrowth measured from eddy covariance,
Bowen ratio, and groundwater flux methods. Decreased saltcedar transpiration
was observed post fire, and locations with higher groundwater levels and lower
salinity (i.e., nearer the stream source) transpired more. Advection of sensible
heat from the nearby desert terrain affected Bowen ratio estimates and especially
in the afternoon at one tower. Since the Bowen ratio method results were
unreliable due to advection, the eddy covariance method was used to truth the
magnitude of the calculated groundwater evapotranspiration. Although there
were many corrections in each method, calibrated saltcedar groundwater
evapotranspiration was shown to be less than short or grass reference conditions
throughout much of the studied Refuge. This benchmark places saltcedar water
use in parallel with native growth; albeit, the correspondent extent of inhabitation
remains a question.
105
5.2 Conclusions
Evapotranspiration Estimation: A Study of Methods in the Western United
States investigated evapotranspiration methods of Penman-Monteith, Bowen
ratio, eddy covariance, groundwater fluxes, and an assimilated dataset compared
to sap flow and remote sensing. While there are further measurement
techniques, these were contrasted at a large and small areal scale to
authenticate the validity of the models in the mountainous and arid American
West. In the comparisons, it was evident that, generally, much greater spatial
and temporal measurements are needed for full analysis. New ideas and better
and more instruments will certainly be introduced, which will begin to fulfill these
requirements. Nonetheless, concerted efforts to capture the detailed water cycle
through greater data acquisition are already within reach but not currently
realized. In short, this deficiency underscores ‘what could have been’ and what
can happen.
5.3 References
Brutsaert, W. 1982. Evaporation into the atmosphere: theory, history, and applications.
GPM. Accessed September 8, 2014. Goddard Space Flight Center | National Aeronautics and Space Administration. <http://www.nasa.gov/centers/goddard/news/gpm_review.html>
Landsat. Accessed September 8, 2014. Landsat Missions | U.S. Geological Survey. <http://landsat.usgs.gov/index.php>
NLDAS-2. The North American Land Data Assimilation System Phase 2. NASA Goddard Space Flight Center and Other Collaborators. <ftp://hydro1.sci.gsfc.nasa.gov/data/s4pa//NLDAS/NLDAS_FORA0125_H.002/>
108
ELSEVIER LICENSE TERMS AND CONDITIONS Sep 04, 2014
This is a License Agreement between Clayton S Lewis ("You") and Elsevier
("Elsevier") provided by Copyright Clearance Center ("CCC"). The license consists
of your order details, the terms and conditions provided by Elsevier, and the
payment terms and conditions.
All payments must be made in full to CCC. For payment instructions, please see information listed at the bottom of this form.
Supplier Elsevier Limited The Boulevard,Langford Lane Kidlington,Oxford,OX5 1GB,UK
Registered Company Number 1982084 Customer name Clayton S Lewis Customer address Civil and Environmental Engineering LOGAN, UT 84321 License number 3462140939180 License date Sep 04, 2014 Licensed content publisher Elsevier Licensed content publication Journal of Hydrology
Licensed content title Comparison of the NLDAS Weather Forcing Model to Agrometeorological Measurements in the western United States
Licensed content author Clayton S. Lewis,Hatim M.E. Geli,Christopher M.U. Neale Licensed content date 14 March 2014 Licensed content volume number 510
Licensed content issue number n/a Number of pages 8 Start Page 385
109
End Page 392 Type of Use reuse in a thesis/dissertation Portion full article Format both print and electronic Are you the author of this Elsevier article? Yes
Will you be translating? No
Title of your thesis/dissertation Evapotranspiration Estimation: A Study of Methods in the Western United States
Expected completion date Dec 2014 Estimated size (number of pages) 125
Elsevier VAT number GB 494 6272 12 Permissions price 0.00 USD VAT/Local Sales Tax 0.00 USD / 0.00 GBP Total 0.00 USD Terms and Conditions
INTRODUCTION
1. The publisher for this copyrighted material is Elsevier. By clicking "accept" in connection with completing this licensing transaction, you agree that the following terms and conditions apply to this transaction (along with the Billing and Payment terms and conditions established by Copyright Clearance Center, Inc. ("CCC"), at the time that you opened your Rightslink account and that are available at any time at http://myaccount.copyright.com).
GENERAL TERMS
2. Elsevier hereby grants you permission to reproduce the aforementioned material subject to the terms and conditions indicated.
3. Acknowledgement: If any part of the material to be used (for example, figures) has appeared in our publication with credit or acknowledgement to another source, permission must also be sought from that source. If such permission is not obtained then that material may not be included in your publication/copies. Suitable acknowledgement to the source must be made, either as a footnote or in a reference list at the end of your publication, as follows:
“Reprinted from Publication title, Vol /edition number, Author(s), Title of article / title of chapter, Pages No., Copyright (Year), with permission from Elsevier [OR APPLICABLE SOCIETY COPYRIGHT OWNER].” Also Lancet special credit - “Reprinted from The Lancet, Vol. number, Author(s), Title of article, Pages No., Copyright (Year), with
110
permission from Elsevier.”
4. Reproduction of this material is confined to the purpose and/or media for which permission is hereby given.
5. Altering/Modifying Material: Not Permitted. However figures and illustrations may be altered/adapted minimally to serve your work. Any other abbreviations, additions, deletions and/or any other alterations shall be made only with prior written authorization of Elsevier Ltd. (Please contact Elsevier at [email protected])
6. If the permission fee for the requested use of our material is waived in this instance, please be advised that your future requests for Elsevier materials may attract a fee.
7. Reservation of Rights: Publisher reserves all rights not specifically granted in the combination of (i) the license details provided by you and accepted in the course of this licensing transaction, (ii) these terms and conditions and (iii) CCC's Billing and Payment terms and conditions.
8. License Contingent Upon Payment: While you may exercise the rights licensed immediately upon issuance of the license at the end of the licensing process for the transaction, provided that you have disclosed complete and accurate details of your proposed use, no license is finally effective unless and until full payment is received from you (either by publisher or by CCC) as provided in CCC's Billing and Payment terms and conditions. If full payment is not received on a timely basis, then any license preliminarily granted shall be deemed automatically revoked and shall be void as if never granted. Further, in the event that you breach any of these terms and conditions or any of CCC's Billing and Payment terms and conditions, the license is automatically revoked and shall be void as if never granted. Use of materials as described in a revoked license, as well as any use of the materials beyond the scope of an unrevoked license, may constitute copyright infringement and publisher reserves the right to take any and all action to protect its copyright in the materials.
9. Warranties: Publisher makes no representations or warranties with respect to the licensed material.
10. Indemnity: You hereby indemnify and agree to hold harmless publisher and CCC, and their respective officers, directors, employees and agents, from and against any and all claims arising out of your use of the licensed material other than as specifically authorized pursuant to this license.
11. No Transfer of License: This license is personal to you and may not be sublicensed, assigned, or transferred by you to any other person without publisher's written permission.
12. No Amendment Except in Writing: This license may not be amended except in a
111
writing signed by both parties (or, in the case of publisher, by CCC on publisher's behalf).
13. Objection to Contrary Terms: Publisher hereby objects to any terms contained in any purchase order, acknowledgment, check endorsement or other writing prepared by you, which terms are inconsistent with these terms and conditions or CCC's Billing and Payment terms and conditions. These terms and conditions, together with CCC's Billing and Payment terms and conditions (which are incorporated herein), comprise the entire agreement between you and publisher (and CCC) concerning this licensing transaction. In the event of any conflict between your obligations established by these terms and conditions and those established by CCC's Billing and Payment terms and conditions, these terms and conditions shall control.
14. Revocation: Elsevier or Copyright Clearance Center may deny the permissions described in this License at their sole discretion, for any reason or no reason, with a full refund payable to you. Notice of such denial will be made using the contact information provided by you. Failure to receive such notice will not alter or invalidate the denial. In no event will Elsevier or Copyright Clearance Center be responsible or liable for any costs, expenses or damage incurred by you as a result of a denial of your permission request, other than a refund of the amount(s) paid by you to Elsevier and/or Copyright Clearance Center for denied permissions.
LIMITED LICENSE
The following terms and conditions apply only to specific license types:
15. Translation: This permission is granted for non-exclusive world English rights only unless your license was granted for translation rights. If you licensed translation rights you may only translate this content into the languages you requested. A professional translator must perform all translations and reproduce the content word for word preserving the integrity of the article. If this license is to re-use 1 or 2 figures then permission is granted for non-exclusive world rights in all languages.
16. Posting licensed content on any Website: The following terms and conditions apply as follows: Licensing material from an Elsevier journal: All content posted to the web site must maintain the copyright information line on the bottom of each image; A hyper-text must be included to the Homepage of the journal from which you are licensing at http://www.sciencedirect.com/science/journal/xxxxx or the Elsevier homepage for books at http://www.elsevier.com; Central Storage: This license does not include permission for a scanned version of the material to be stored in a central repository such as that provided by Heron/XanEdu.
Licensing material from an Elsevier book: A hyper-text link must be included to the Elsevier homepage at http://www.elsevier.com . All content posted to the web site must maintain the copyright information line on the bottom of each image.
112
Posting licensed content on Electronic reserve: In addition to the above the following clauses are applicable: The web site must be password-protected and made available only to bona fide students registered on a relevant course. This permission is granted for 1 year only. You may obtain a new license for future website posting.
For journal authors: the following clauses are applicable in addition to the above: Permission granted is limited to the author accepted manuscript version* of your paper.
*Accepted Author Manuscript (AAM) Definition: An accepted author manuscript (AAM) is the author’s version of the manuscript of an article that has been accepted for publication and which may include any author-incorporated changes suggested through the processes of submission processing, peer review, and editor-author communications. AAMs do not include other publisher value-added contributions such as copy-editing, formatting, technical enhancements and (if relevant) pagination.
You are not allowed to download and post the published journal article (whether PDF or HTML, proof or final version), nor may you scan the printed edition to create an electronic version. A hyper-text must be included to the Homepage of the journal from which you are licensing at http://www.sciencedirect.com/science/journal/xxxxx. As part of our normal production process, you will receive an e-mail notice when your article appears on Elsevier’s online service ScienceDirect (www.sciencedirect.com). That e-mail will include the article’s Digital Object Identifier (DOI). This number provides the electronic link to the published article and should be included in the posting of your personal version. We ask that you wait until you receive this e-mail and have the DOI to do any posting.
Posting to a repository: Authors may post their AAM immediately to their employer’s institutional repository for internal use only and may make their manuscript publically available after the journal-specific embargo period has ended.
Please also refer to Elsevier's Article Posting Policy for further information.
18. For book authors the following clauses are applicable in addition to the above: Authors are permitted to place a brief summary of their work online only.. You are not allowed to download and post the published electronic version of your chapter, nor may you scan the printed edition to create an electronic version. Posting to a repository: Authors are permitted to post a summary of their chapter only in their institution’s repository.
20. Thesis/Dissertation: If your license is for use in a thesis/dissertation your thesis may be submitted to your institution in either print or electronic form. Should your thesis be published commercially, please reapply for permission. These requirements include permission for the Library and Archives of Canada to supply single copies, on demand, of the complete thesis and include permission for UMI to supply single copies, on demand, of
113
the complete thesis. Should your thesis be published commercially, please reapply for permission.
Elsevier Open Access Terms and Conditions
Elsevier publishes Open Access articles in both its Open Access journals and via its Open Access articles option in subscription journals.
Authors publishing in an Open Access journal or who choose to make their article Open Access in an Elsevier subscription journal select one of the following Creative Commons user licenses, which define how a reader may reuse their work: Creative Commons Attribution License (CC BY), Creative Commons Attribution – Non Commercial - ShareAlike (CC BY NC SA) and Creative Commons Attribution – Non Commercial – No Derivatives (CC BY NC ND)
Terms & Conditions applicable to all Elsevier Open Access articles:
Any reuse of the article must not represent the author as endorsing the adaptation of the article nor should the article be modified in such a way as to damage the author’s honour or reputation.
The author(s) must be appropriately credited.
If any part of the material to be used (for example, figures) has appeared in our publication with credit or acknowledgement to another source it is the responsibility of the user to ensure their reuse complies with the terms and conditions determined by the rights holder.
Additional Terms & Conditions applicable to each Creative Commons user license:
CC BY: You may distribute and copy the article, create extracts, abstracts, and other revised versions, adaptations or derivative works of or from an article (such as a translation), to include in a collective work (such as an anthology), to text or data mine the article, including for commercial purposes without permission from Elsevier
CC BY NC SA: For non-commercial purposes you may distribute and copy the article, create extracts, abstracts and other revised versions, adaptations or derivative works of or from an article (such as a translation), to include in a collective work (such as an anthology), to text and data mine the article and license new adaptations or creations under identical terms without permission from Elsevier
CC BY NC ND: For non-commercial purposes you may distribute and copy the article and include it in a collective work (such as an anthology), provided you do not alter or modify
114
the article, without permission from Elsevier
Any commercial reuse of Open Access articles published with a CC BY NC SA or CC BY NC ND license requires permission from Elsevier and will be subject to a fee.
Commercial reuse includes:
• Promotional purposes (advertising or marketing)
• Commercial exploitation ( e.g. a product for sale or loan)
• Systematic distribution (for a fee or free of charge)
Please refer to Elsevier's Open Access Policy for further information.
21. Other Conditions:
v1.6 You will be invoiced within 48 hours of this transaction date. You may pay your invoice by credit card upon receipt of the invoice for this transaction. Please follow instructions provided at that time. To pay for this transaction now; please remit a copy of this document along with your payment. Payment should be in the form of a check or money order referencing your account number and this invoice number RLNK501395028. Make payments to "COPYRIGHT CLEARANCE CENTER" and send to: Copyright Clearance Center Dept 001 P.O. Box 843006 Boston, MA 02284-3006 Please disregard electronic and mailed copies if you remit payment in advance. Questions? [email protected] or +1-855-239-3415 (toll free in the US) or +1-978-646-2777.
Gratis licenses (referencing $0 in the Total field) are free. Please retain this printable license for your reference. No payment is required.
117
CURRICULUM VITAE
CLAYTON SCOTT LEWIS
Address
Office: Utah State University Civil and Environmental Engineering Engineering Innovation 206A 4110 Old Main Hill Logan, Utah 84322-4110
Home: 537 East 300 North Logan, Utah 84321-6448
Phone
Office: (435) 797-1041
Academic: [email protected] INTERESTS Integrated solutions to complex, real world problems involving storage, processing, and transfers of huge datasets with scientific, mathematical, and spatial properties. Professional licensure. Spending time with family. EDUCATION Doctor of Philosophy Degree in Civil Engineering, Water Emphasis, (2014) Evapotranspiration Estimation: A Study of Methods in the Western United States Utah State University, Logan, Utah, USA
Master of Engineering Degree in Civil Engineering, Structural Emphasis, 2012 Utah State University, Logan, Utah, USA
Bachelor of Science Degree in Civil Engineering, 2011 Utah State University, Logan, Utah, USA
Passed Fundamentals of Engineering Exam, 2010 EMPLOYMENT
Research Assistant (2008-Present) • Developed custom research software for several water resources
engineering projects utilizing the .NET framework, ArcGIS environment,
118
SQL, and free libraries for local, state, and federal government agencies, a sampling: Determined consumptive use for the state of Utah for more than 20
land covers at 246 locations spanning a 38 year period to regulate water rights applications
Established landscape irrigation efficiencies from satellite imagery, evapotranspiration data, and property water use for municipalities and conservancy districts in Utah, which included reports and emails to customers
Calculated instantaneous actual evapotranspiration using the two source energy balance model with variable input from satellite, aerial, and point source data
Analyzed an hourly weather dataset of 704 weather stations over a 32 year period to validate a climate model and establish agricultural water use in 17 western states
Investigated potential improvements in pump scheduling, pipe capacity, and peak water demands in Moldova and Lebanon
• Delivered more than 20 presentations, many of which were marketing a software in Utah and Colorado to municipalities and water conservancy districts
• Collaboratively wrote/contributed to five major reports and others on a smaller scale
• Maintained and managed a network of over 30 electronic weather stations across the state of Utah and in California
PUBLICATIONS Lewis, C. S., Geli, H. M., & Neale, C. M. (2014). Comparison of the NLDAS Weather Forcing Model to Agrometeorological Measurements in the Western United States. Journal of Hydrology, 510, 385-392. Hill, R. W., Barker, J. B., Lewis, C. S. (2011). Crop and Wetland Consumptive Use and Open Water Surface Evaporation for Utah. Utah Agricultural Experiment Station Research Report, 213. PRESENTATIONS (Audience) City of Gillette, Wyoming City of North Salt Lake, Utah City of Thornton, Colorado Colorado Springs, Colorado Jordan Valley Water Conservancy District Logan City, Utah Multiple Municipalities in the Salt Lake County Area, Utah Multiple Municipalities in the Utah County Area, Utah Northern Colorado Water Conservancy District
119
Salt Lake City, Utah Six Agencies Committee of the Lower Colorado River (Webinar) Utah Divisions of Water Resources and Water Rights Utah State University Interdepartmental Utah State University Spring Runoff Conference Weber Basin Water Conservancy District Western Water Assessment (Webinar) SKILLSETS Computer Visual Studio, Visual Basic.NET (C#), Git, SQL (Microsoft SQL Server, SQLite, etc.), Networking, Report Generation, Campbell Scientific Dataloggers, GDAL, ArcGIS, MapWindow, RISA, SAP, AutoCAD, MATLAB, Maple, Excel, Word, Powerpoint, Access, Open Office, Dabbling in Python, Java, C++, … Languages English (Native), Russian