effects of snowpack parameters and layering processes at x- and ku-band backscatter ali nadir arslan...

Effects of Snowpack Parameters

and Layering Processes

at X- and Ku-band Backscatter

Ali Nadir Arslan1, Jouni Pulliainen1, Juha Lemmetyinen1,

Thomas Nagler 2,Helmut Rott2, Michael Kern3

Finnish Meteorological Institute, Erik Palménin aukio 1, P.O.Box 503, FI-00101 Helsinki, Finland

Email: [email protected]

2ENVEO IT GmbH ICT Technologiepark Technikerstrasse 21a, 6020 Innsbruck,Austria

3European Space Research and Technology Centre (ESTEC) Keplerlaan 1 2200 AG Noordwijk, The

Netherlands

2

BACKGROUND:• The Cold Regions Hydrology High-Resolution Observatory

(CoReH2O) mission has been selected as one of the candidate Earth Explorer missions for studies of technical and scientific feasibility and consolidation at Phase A level.

• The CoReH2O mission aims to close gaps in spatially detailed observations of key parameters of the global snow and ice masses for applications in climate research and hydrology.

• The proposed technical solution relies on a dual frequency, X- and Ku-band (9.6 and 17.2 GHz), dual-polarized (VV and VH) synthetic aperture radar. This study is conducted under the development of the geophysical algorithm of CoReH2O, a proposed ESA SAR mission currently in Phase A .

3

ACTIVITIES:• Sensitivity study on X- and Ku-band backscattering for different snow

physical states (taking into accounts effects of snow grain size, metamorphic state, layering, melting state) using theoretical forward models and experimental data.

• Sensitivity study on X- and Ku-band backscattering for effects of local incidence angle and terrain slope taking into account different snow morphologies, including the assessment of uncertainties in the DEM used to compute the local incidence angle and slope.

• Sensitivity study on X-and Ku-band backscattering for effects of different soil properties (surface roughness, wetness, freezing conditions).

• Sensitivity study on X-and Ku-band backscattering of winter snow accumulation on glaciers and impact of different background conditions of glaciers surfaces (the glacier facies).

• Assessment and documentation of the impact of these effects for the retrieval of SWE and snow depth, including suggestions for procedures and auxiliary data sets to compensate for these effects.

4

ESA Nordic Snow Radar Experiment (NoSREx)Reference instruments for space-borne monitoring of the cryosphere

Goal: Provide Data for other ongoing CoReH20 studies:

(1) CoReH2O geophysical algorithm development

(2) studies of the synergistic use of CoReH2O measurements with other (passive) observational data

Means:

• Experimental dataset on the backscattering and brightness temperature characteristics of snow

• Winter seasons of 2009-2010 and 2010-2011

• Active X to Ku band microwave observations with ESA SnowScat-system

• Passive mw observations at L-, X-, K-, Ka- and W- band radiometers

• in situ measurements of snow cover, soil and atmospheric properties

520.04.23 5

Intensive Observation Area (IOA)

6721.712 N 2638.270 E

Site typical boreal coniferous forest on mineral soil

Average permanent snow cover: 6th Nov – 25 May (1971-2000)

Average maximum snow depth: 80 cm

Easy access and technical support

Sodankylä

FMI Arctic Research Centre, Sodankylä, Finland

6

Measurement towers for instrument installation (5 m, 8m, 38m)

In vicinity of meteorological/ atmospheric sounding observations and CO2 flux measurements

Manual snow cover measurements on site

Automatic sensors (soil moisture and temperature profile, SWE, snow depth, snow temperature profile)

L to W band radiometers

Bi- weekly snowpits

X- to Ku-band scatterometer

Automatic sensors (Soil moisture, Temperature, bulk SWE, Snow Depth)

Photo: webcam on 38 m tower Intensive Observation Area (IOA)

7

Active microwave observations - SnowScat

Parameter Nominal Observations Sequence start 01:45 (UTC) + every 3 hrs Sequence total duration 1 h 45 minutes Integration time 0.00256 s Azimuth scan range*) -62 < < 34 Azimuth scan step 6 Elevation scan range 30 < < 60 Elevation scan step 10 Frequency 9.199984 – 17.798512 GHz Polarization HH,HV,VH,VV Step duration 45 s Estimated data per sequence 8 MB

Internal calibration loop Before each polarization cycle (at every measurement angle)

Calibration target Before and after each measurement cycle

• Frequency scanning scatterometer, stepped CW from 9.15 to 17.9 GHz

• Manufacturer GAMMA remote sensing, Instrument on loan from ESA

• measurement every 3 h

• Azimuth scan of 100 (6 steps, 17 looks)

• Data available at inc. angles 30, 40, 50, 60• HH/VV, HV/VH

• Every measurement includes two views of calibration sphere

8

MODELING:• The classical DMRT approach (e.g. Shi; 2006, Du et al., 2007) was

selected due the simplicity although the selected model ignores the coherent wave near-field interaction among the particles which results in underestimation of depolarization effects and overestimation of frequency dependence.

• Shi, J. 2006. Snow water equivalence retrieval using X and Ku band dual-polarization radar. Proceedings of IGARSS 2006, 2183-2185.

• Du, J., J. Shi, S. Tjuatja, and K.S. Chen. 2007. A multi-scattering and multi-layer snow model and its validation. Proceedings of IGARSS2007, 1219-1222.

• A multilayer-snow electromagnetic (EM) backscattering model based on the vector radiative transfer, the strong fluctuation theory was developed and compared with DMRT and measurement data.

9

The main input parameters for the second-order DMRT model

10

CONLUSION REMARK :• Grain size dependence of the backscatter is critical for SWE

retrieval and rather similar at co-and cross-polarized channels at both X-and Ku-band.

• Backscatter coefficient in dependence of incidence angle at X-(left) and Ku-band (right), snow depth = 1m, snow density=250kg /m3, rts=0.25, rad=0.1, 0.25, 0.5, 0.75, 1mm, frozen soil.

X-band, VV-polarization

-25

-20

-15

-10

-5

0

10 20 30 40 50 60

Incidence angle (deg)

Ba

cks

catt

erin

g c

oe

ffic

ien

t (d

B)

radius = 0,1 mm

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

radius = 1 mm

X-band, HH-polarization

-25

-20

-15

-10

-5

0

10 20 30 40 50 60


Ba

cksc

atte

rin

g C

oef

fici

en

t (d

B)

radius = 0,1 mm

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

radius = 1 mm

X-band, VH-polarization

-40

-35

-30

-25

-20

-15

-10

-5

0

10 20 30 40 50 60


Ba

cksc

atte

rin

g c

oe

ffic

ien

t (d

B)

radius = 0,1 mm

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

radius = 1 mm

Ku-band, VV-polarization

-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

10 20 30 40 50 60


Ba

ck

sc

att

eri

ng

co

eff

icie

nt

(dB

)

radius = 0,1 mm

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

radius = 1 mm

Ku-band, HH-polarization

-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

10 20 30 40 50 60


Ba

ck

sc

att

eri

ng

co

eff

icie

nt

(dB

)

radius = 0,1 mm

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

radius = 1 mm

Ku-band, VH-polarization

-35

-30

-25

-20

-15

-10

-5

0

10 20 30 40 50 60


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,1 mm

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

radius = 1 mm

X-bandHH-Pol.

Ku-bandHH-Pol.

13

CONLUSION REMARK:• The dependence of backscatter on ratio of snow grain

short/long axis (rts) is of small effect for co-polarized backscatter, but more effect on cross-polarized backscatter for both X-and Ku-band.

• Backscatter coefficient in dependence of incidence angle at X-(left) and Ku-band (right), snow depth = 1m, snow density=250kg /m3, rts=0.25, 0.50, 0.75 and 1.0 rad=0.5mm, frozen soil.


-12

-10

-8

-6

-4

-2

0

10 20 30 40 50 60


Ba

cksc

atte

rin

g c

oef

fici

ent (

dB

)

RTS=0,25

RTS=0,50

RTS=0,75

RTS=1,0


-12

-10

-8

-6

-4

-2

0

10 20 30 40 50 60


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

RTS=0,25

RTS=0,50

RTS=0,75

RTS=1,0


-30

-25

-20

-15

-10

-5

0

10 20 30 40 50 60


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

RTS=0,25

RTS=0,50

RTS=0,75

RTS=1,0


-18

-16

-14

-12

-10

-8

-6

-4

-2

0

10 20 30 40 50 60


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

RTS=0,25

RTS=0,50

RTS=0,75

RTS=1,0


-18

-16

-14

-12

-10

-8

-6

-4

-2

0

10 20 30 40 50 60


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

RTS=0,25

RTS=0,50

RTS=0,75

RTS=1,0


-40

-35

-30

-25

-20

-15

-10

-5

0

10 20 30 40 50 60


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

RTS=0,25

RTS=0,50

RTS=0,75

RTS=1,0

15

CONLUSION REMARK:• The sensitivity of backscatter to increase in SWE mostly

depends on grain size. Ku-band is more sensitive to change in SWE than X-band.

• Backscatter coefficient in dependence of SWE at X-(left) and Ku-band (right), snow density= 250kg /m3 , rts=0.25, rad=0.25, 0.50, 0.75;background is frozen soil with soilrms = 3mm, incidence angle 40 degrees .


-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400

Snow water equivalent (mm)

Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

X-bandHH-Pol.

Ku-bandHH-Pol.

18

CONLUSION REMARK:• The snow density dependence of the backscatter is not

critical for SWE retrieval.

• Backscatter coefficient in dependence of SWE at X-(left) and Ku-band (right), snow density= 300kg /m3 , rts=0.25, rad=0.25, 0.50, 0.75;background is frozen soil with soilrms = 3mm, incidence angle 40 degrees .


-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-18

-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400

Snow water equivalent (dB)

Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm


-40

-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

radius = 0,25 mm

radius = 0,5 mm

radius = 0,75 mm

20

CONLUSION REMARK:• The layering of snow pack changes the sensitivity of

backscatter to SWE. A layer of refrozen at the bottom of snow pack (resulting from thaw-refreeze cycles at early winter) can cause a negative correlation of backscatter with the increase SWE for the beginning of the dry snow accumulation period.

21

• We extracted and analyzed a 30-year-long data set of snow depth and air temperature for the Sodankylä station, northern Finland. Sodankylä can be considered as a quite typical representative of the Eurasian and North American continental boreal/sub-arctic climate, especially with respect to snow cover conditions (Sturm et al., 1995). The main difference with respect to more central continental conditions is that thaw/refreeze event during mid-winter are more frequent at Sodankylä that is for some winters influenced by North Atlantic low pressure systems transporting warmer air (resulting to thaw-refreeze events even during the mid-winter).

• Based on Sodankylä time-series conditions with thaw/refreeze events are very typical during the early winter, but not so usual later on winter due to general decrease of air temperature towards the mid-winter. This results to snow pack structure with large refrozen snow grains on the bottom of the snow pack (Kontu and Pulliainen, 2010; Lemmetyinen et al., 2010) . Another process that produces snow packs with large grains on the bottom is the generation of depth hoar, which is typical both in North America and northern Eurasia, e.g. Rees et al. (2010) have reported observations on typical grain size profiles for the Canadian tundra.

• In case of Sodankylä, the 30-year time-series of weather station data suggests that in 73% of all winters, a layer of refrozen snow exists below the height of 20 cm , and in 57% of all cases this layer is certainly locating at the height below 10 cm. Thus, we investigate here how the layering of snow pack changes the sensitivity when compared with the single layer case by inputting a dense snow layer with coarse grain-sized snow below with fine grain-sized snow above (with varying SWE and increasing snow grain size as a parameter). As discussed above, this corresponds to a typical situation where refrozen snow from early winter is below a newer mid-winter fallen snow layer (and also the case of depth hoar).

22

Snow densities of layer 1 (rho1) and layer 2 (rho2) were set to 160 and to 300 kg/m3 , respectively. Incidence angle (theta) was set to 40 degrees . Snowrms and soilrms were set to 2mm and 3mmrespectively. The depth of snow layer 2 (d2) was kept as 10 cm for all simulations and the depthof snow layer 1 (d1) was increased from 20cm to 200cm. The grain size of snow layer 2 waskept constant as rad2 = 1mm and the grain size of snow layer 1 was set as rad1 = 0.1mm, 0.2mm, 0.3mm, 0.4mm, 0.5mm, 0.6mm.

24

X-bandHH-Pol.

Ku-bandHH-Pol.

25

CONLUSION REMARK:• The positive correlation between snow grain size and SWE,

typical for the temporal metamorphosis, increases the correlation between SWE and backscattering coefficient.

Temporal evolution of snow pack through a typical metamorphosis process where depth hoar is evolving and snow grain size in the top layer increases with time.

X-band

-35

-30

-25

-20

-15

-10

-5

0

125 150 175 200 225 250 275


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

VV

HH

VH

Ku-band

-30

-25

-20

-15

-10

-5

0

125 150 175 200 225 250 275


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

VV

HH

VH

27

CONLUSION REMARK:• Surface roughness characteristics of the underlaying terrain

have a significant effect to the sensitivity of backscatter to variations in SWE.


-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

soil_rms = 1mm

soil_rms=3mm

soil_rms=5mm


-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

soil_rms = 1mm

soil_rms=3mm

soil_rms=5mm


-40

-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

soil_rms = 1mm

soil_rms=3mm

soil_rms=5mm


-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

soil_rms = 1mm

soil_rms=3mm

soil_rms=5mm


-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

soil_rms = 1mm

soil_rms=3mm

soil_rms=5mm


-50

-40

-30

-20

-10

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

soil_rms = 1mm

soil_rms=3mm

soil_rms=5mm


-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

wet soil

frozen soil


-16

-14

-12

-10

-8

-6

-4

-2

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

wet soil

frozen soil


-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

wet soil

frozen soil


-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

wet soil

frozen soil


-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

wet soil

frozen soil


-35

-30

-25

-20

-15

-10

-5

0

25 50 75 100 150 200 250 300 350 400


Bac

ksca

tter

ing

co

effi

cien

t (d

B)

wet soil

frozen soil

30

CONLUSION REMARKS:• Effect of grain size is much bigger than SWE; retrieval

algorithms have to compensate the effect of grain size.

31

Comparison with the measurement data:

X-bandHH-Pol.

Ku-bandHH-Pol.

32

Sigma0 (dB)

Typical value

Grain size

Max grain size 1 mm

Min grain size 0.1 mm

…..

Grain shape ….. All min/max effects

33

Set of parameters:• Use frequencies: X, Ku, VV and VH

• Proposed range of parameters for computing sigma0, for given values of SWE (25mm, 50mm, 75mm, 100mm, 150mm 200mm, 250mm, 300mm, 350mm, 400mm):

• Grain RADIUS: 0.25 mm, 0.50 mm, 0.75mm

• Grain shape: sphere, ellipsoid b/a = 0.25

• Incidince angle: 30 degrees, 40 degrees, 50 degrees

• Snow density (mean values of snowpack) 250 kg/m3; 300 kg/m3

• Snow temperature: -5°c

• Background target: soil std dev. Height 1 mm, 3mm,5 mm; state: frozen, wet (water content mv = 0.10)

34

X-band, VV

-2

0

2

4

6

8

10

12

grainsize soilrms snowdensity soil dry/wet incidence_angle

Different Parameters

Sig

ma0

(d

B)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

X-band, VV

-2

0

2

4

6

8

10

12

SWE

Sig

ma0

(d

B) grainsize

soilrms

snowdensity

soil dry/wet

incidence_angle

35

X-band, VH

-2

0

2

4

6

8

10

12

14

16



Sig

ma0

(d

B)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

X-band, VH

-2

0

2

4

6

8

10

12

14

16

SWE

Sig

ma0

(d

B) grainsize

soilrms

snowdensity

soil dry/wet

incidence_angle

36

Ku-band, VV

-2

0

2

4

6

8

10

12



Sig

ma0

(d

B)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

Ku-band, VV

-2

0

2

4

6

8

10

12

SWE

Sig

ma0

(d

B) grainsize

soilrms

snowdensity

soil dry/wet

incidence_angle

37

Ku-band, VH

-2

0

2

4

6

8

10

12

14



Sig

ma0

(d

B)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

Ku-band, VH

-2

0

2

4

6

8

10

12

14

swe

=25m

m

swe

=50m

m

swe

=75m

m

swe

=100

mm

swe

=150

mm

swe

=200

mm

swe

=250

mm

swe

=300

mm

swe

=350

mm

swe

=400

mm

SWE

Sig

ma0

(d

B) grainsize

soilrms

snowdensity

soil dry/wet

incidence_angle

38

X-band, VV

-2

0

2

4

6

8

10

12


Sig

ma0

(dB

)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

X-band, VH

-202468

10121416


Sig

ma0

(dB

)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

Ku-band, VV

-2

0

2

4

6

8

10

12


Sig

ma0

(dB

)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

Ku-band, VH

-2

0

2

4

6

8

10

12

14


Sig

ma0

(dB

)

swe =25mm

swe =50mm

swe =75mm

swe =100mm

swe =150mm

swe =200mm

swe =250mm

swe =300mm

swe =350mm

swe =400mm

39

Soil RMS differences effect on Sigma0

X-band, VV

0

2

4

6

8

10

12

swe

=25m

m

swe

=50m

m

swe

=75m

m

swe

=100

mm

swe

=150

mm

swe

=200

mm

swe

=250

mm

swe

=300

mm

swe

=350

mm

swe

=400

mm

SWE

Sig

ma0

(d

B)

1-5mm

1-3mm

3-5mm

X-band, VH

0

24

6

8

1012

14

16

swe

=25m

m

swe

=50m

m

swe

=75m

m

swe

=100

mm

swe

=150

mm

swe

=200

mm

swe

=250

mm

swe

=300

mm

swe

=350

mm

swe

=400

mm

SWE

Sig

ma0

(d

B)

1-5mm

1-3mm

3-5mm

Ku-band, VV

0

12

3

4

56

7

8

swe

=25m

m

swe

=50m

m

swe

=75m

m

swe

=100

mm

swe

=150

mm

swe

=200

mm

swe

=250

mm

swe

=300

mm

swe

=350

mm

swe

=400

mm

SWE

Sig

ma0

(d

B)

1-5mm

1-3mm

3-5mm

Ku-band, VH

0

2

4

6

8

10

12

swe

=25m

m

swe

=50m

m

swe

=75m

m

swe

=100

mm

swe

=150

mm

swe

=200

mm

swe

=250

mm

swe

=300

mm

swe

=350

mm

swe

=400

mm

SWE

Sig

ma0

(d

B)

1-5mm

1-3mm

3-5mm

40

KEY FINDINGS:• Grain size dependence of the backscatter is critical for SWE retrieval and rather

similar at co-and cross-polarized channels at both X-and Ku-band.• The dependence of backscatter on ratio of snow grain short/long axis (rts) is of

small effect for co-polarized backscatter, but more effect on cross-polarized backscatter for both X-and Ku-band.

• The sensitivity of backscatter to increase in SWE mostly depends on grain size. Ku-band is more sensitive to change in SWE than X-band.

• The snow density dependence of the backscatter is not critical for SWE retrieval.

• The layering of snow pack changes the sensitivity of backscatter to SWE. A layer of refrozen at the bottom of snow pack (resulting from thaw-refreeze cycles at early winter) can cause a negative correlation of backscatter with the increase SWE for the beginning of the dry snow accumulation period. The positive correlation between snow grain size and SWE, typical for the temporal metamorphosis, increases the correlation between SWE and backscattering coefficient.

• Surface roughness characteristics of the underlaying terrain have a significant effect to the sensitivity of backscatter to variations in SWE.

• Effect of grain size is much bigger than SWE; retrieval algorithms have to compensate the effect of grain size.

effects of snowpack parameters and layering processes at x- and ku-band backscatter ali nadir arslan...

Documents