october 11 15 2010 hamburg germany october 11-15, 2010...
TRANSCRIPT
October 11 15 2010 Hamburg Germany
1(*) 1 1 2 3 4
October 11-15, 2010 Hamburg, Germany
Emad Habib1(*), Alemseged T. Haile1, Mohamed Elsaadani1, Mohamed ElShamy2, Robert Kuligowski3, & Yudong Tian4Emad Habib , Alemseged T. Haile , Mohamed Elsaadani , Mohamed ElShamy , Robert Kuligowski , & Yudong Tian1Dept Civil Engineering University of Louisiana at Lafayette LA USA; 2Nile Forecasting Center Cairo Egypt; 3NOAA/NESDIS/CSTAR; 4NASA/GSFC1Dept. Civil Engineering, University of Louisiana at Lafayette, LA, USA; 2Nile Forecasting Center, Cairo, Egypt; 3NOAA/NESDIS/CSTAR; 4NASA/GSFC
** Corresponding Author, [email protected], ph.: 337-482-6513p g , @ , p
Diurnal cycle of rainfall in Lake Victoria basin source of White Nile• This study reports on validation of two high-resolution satellite rainfall products: TRMM 3B42 3-hourly Diurnal cycle of rainfall in Lake Victoria basin, source of White Nile0.25° product, and CMORPH 8-km 30-minute product.
Analysis period: April 1998 2008• The two products are assessed in terms of their ability to re-produce spatial variability in monthly and 1.25 Analysis period: April 1998-2008 (TRMM-3B42) April 2003-2008
p y p p y ydiurnal rainfall cycles across the full Nile Basin domain in Africa.
1
CMORPH
TRMM-3B42 (TRMM 3B42), April 2003 2008 (CMORPH), and April 1969–1976
y
• The CMORPH product is also used to drive a basin-wide hydrological forecasting model (Nile Forecasting1
- 1)
TRMM 3B42Gauge
( ), p(gauge).
• The CMORPH product is also used to drive a basin-wide hydrological forecasting model (Nile Forecasting System) and compared to simulations based on IR-only rainfall product operationally used by the Nile
0.75
(mm
h
System) and compared to simulations based on IR-only rainfall product operationally used by the Nile Forecasting Center in Egypt
0.5
infa
ll (
Forecasting Center in Egypt.
Fi ll th CMORPH d t i l t d t it fi t l ti i d i d d t i0.25
Rai
• Finally, the CMORPH product is evaluated at its finest resolution using a dense independent rain gauge t k i th L i i i th US f i h d l i ll l t lid ti tt ib t 0
1.25 CMORPH
network in south Louisiana in the US, focusing on hydrologically-relevant validation attributes. 0 3 6 9 12 15 18 21 240
Local time (h)
1
)
TRMM-3B42
Gauge
Seasonal cycle of rainfall across the Nile basin: 1 25
( )
0.75
mm
h- 1
Seasonal cycle of rainfall across the Nile basin: 1.25
CMORPH
TRMM-3B42 0.5fall
(m
comparison against long-term rainfall climatology1
-1)
TRMM-3B42Gauge
0 2
0.5
Rai
nf
comparison against long-term rainfall climatology0.75
mm
h-
0.25
0.5
nfa
ll (
m
0 3 6 9 12 15 18 21 240
Box plots are based on GHCN-monthly rain 0.25
Rai Local time (h)
600
Monthly RainfallBox plots are based on GHCN monthly rain gauge (1950-present). Synmbols on the left
0.25
600
Monthly Rainfall
600 Monthly Rainfall
MeanGondar Ethiopia
g g ( ) yof each box refer to TRMM-3B42 (1998- 0 3 6 9 12 15 18 21 24
0
Local time (h)
1 25
600 Monthly Rainfall
Mean500
Gondar, Ethiopia
2008) ; symbols on right refer to CMORPH (2003 2008)
Local time (h) 1.25
CMORPH500
Khartoum, Sudan400
mm
)
(2003-2008).2.5
CMORPH
1
1)
TRMM-3B42Gauge400
mm
)
,
300fall
(m
2
CMORPH
TRMM-3B42Gauge
0.75
mm
h- 1
300fall
(m
200
Rai
nf
1 5m h
-1) Gauge
0.5
nfa
ll (
m
200
Rai
nf 200
1.5
ll (
mm
0 25
Rai
n200100
1
Rai
nfa
l 0.25100
1 2 3 4 5 6 7 8 9 10 11 120
0.5
R
0 3 6 9 12 15 18 21 240
L l ti (h)
1 2 3 4 5 6 7 8 9 10 11 120
1 2 3 4 5 6 7 8 9 10 11 12Month
0 3 6 9 12 15 18 21 240
Local time (h)1 2 3 4 5 6 7 8 9 10 11 12Month
600
Monthly Rainfall 0 3 6 9 12 15 18 21 24Local time (h)
600 Monthly Rainfall
Mean
Nekemte Ethiopia
Evaluation of satellite rainfall for streamflow simulations in500
Nekemte, Ethiopia
Evaluation of satellite rainfall for streamflow simulations in th Nil B i
600
Monthly Rainfall
Mean400
mm
)
the Nile Basin500
Mean
Malakal, Sudan 300
nfa
ll (
m
x 104
x 104
400) 200
Rai
n
3
3.5x 10
0
3
3.5x 10
0400
(m
m) 200
2.5
3
3 s-1
) 202.5
320
ay-1
)
Observed
Si l t d
300
ain
fall
100
1 5
2
rge
(m 40
1 5
2 40
mm
daSimulated
Dongala(IR) Dongala
IR-onlyIR G
200
Ra
1 2 3 4 5 6 7 8 9 10 11 120
1
1.5
isch
ar 60
801
1.5 60
80 MA
P (
m(IR) Dongala(IR+Gauge)
Streamflow simulations performed using theIR+Gauge
100
1 2 3 4 5 6 7 8 9 10 11 12Month
0
0.5
D 80
100 0
0.580
100100
M Streamflow simulations performed using the Nile Forecasting System (NFS) model for0 600
Monthly Rainfall0 60 120 180 240 300 360
0 1000 60 120 180 240 300 360
0 100100
3 5x 10
4
3 5x 10
4
00
Nile Forecasting System (NFS) model for 2007. Both the IR-based and CMORPH rainfall
1 2 3 4 5 6 7 8 9 10 11 120
Month
600 Monthly Rainfall
MeanGambela, Ethiopia
3
3.5
)
3
3.5 0
20
0
20
2007. Both the IR based and CMORPH rainfall were adjusted using the same monthly weights
Month500
2
2.5
m3 s
-1)
2
2.520
40 day
-1)20
40fixed by NFS.
400
mm
)
1.5
2
arg
e (m
1.5
2 40
60 (mm
d40
60DongalaD l
CMORPHGCMORPH
600 Monthly Rainfall
Mean 300
nfa
ll (
m
0 5
1
Dis
cha
1
60
80 MA
P 60
80
(CMORPH) Dongala(CMORPH+Gauge)+GaugeCMORPH
500
MeanJuba, Sudan
200
Rai
n
0 60 120 180 240 300 3600
0.5
0
0.5
80
100100400m)
200
0 60 120 180 240 300 360Julian Day
0 60 120 180 240 300 360Julian Day
400
ll (
mm 100
12000 0 12000 0
15000 0
15000 0
300
Rai
nfa
l
1 2 3 4 5 6 7 8 9 10 11 120
8000
10000
3 s-1
) 20
Observed 8000
10000 20
y-1
)
15000
) 20
15000
20200
R
Month
6000
8000
ge
(m3
40Observed
Simulated6000
800040
mm
day
Diem
Diem (IR)10000
m3 s
-1)
40Ob d
10000 40 day
-1)Khartoum
(IR)
IR G
IR-only100
600 Monthly Rainfall
MeanJi j U d4000
sch
arg
604000
60
AP
(mDiem
(IR+Gauge)
arg
e (m
60
Observed
Simulated60 (m
m d
Khartoum(IR+Gauge)IR+Gauge IR+Gauge
IR-only0500
MeanJinja, Uganda
2000Dis
80 2000 80 MA
5000
Dis
cha 60
80
500060
80
MA
P y
1 2 3 4 5 6 7 8 9 10 11 120
Month
400m)
0 60 120 180 240 300 3600 100
0 60 120 180 240 300 3600 100
12000 0 12000 0
80 80400
ll (
mm
10000
12000
)
0
2010000
12000 0
20
0 60 120 180 240 300 3600 100
0 60 120 180 240 300 3600 100
600
Monthly Rainfall
300
Rai
nfa
l
8000
m3 s
-1) 20
408000
20
40 day
-1)
15000 0
15000 0
CMORPH500
MeanKitega, Burundi200
R
6000
arg
e (m 40
60
600040
60 (mm
d
Diem(CMORPH+Gauge)
Diem(CMORPH)
s-1) 20 20
y-1
)CMORPH
CMORPH+Gauge
500
100
2000
4000
Dis
cha 60
80 2000
400060
80 MA
P ((CMORPH Gauge)
(CMORPH)10000
ge
(m3
40 10000 40
m d
ay
Khartoum
CMORPHCMORPH
G
400
mm
)
0
0
2000D 80
100 0
2000 80
1005000sch
arg
605000
60
AP
(mKhartoum
(CMORPH)(CMORPH+Gauge)CMORPH +Gauge
300
nfa
ll ( 1 2 3 4 5 6 7 8 9 10 11 12
0
Month
600
Monthly Rainfall
0 60 120 180 240 300 3600
Julian Day
1000 60 120 180 240 300 360
0
Julian Day
100
Dis
80
5000
80
MA
200
Rai
500
MeanKisumu, Kenya
0 60 120 180 240 300 3600 100
0 60 120 180 240 300 3600 100
100
200 500
0 60 120 180 240 300 360Julian Day
0 60 120 180 240 300 360Julian Day
100 400
(mm
)
f1 2 3 4 5 6 7 8 9 10 11 12
0 300
infa
ll (
Evaluation at fine time-space resolution using dense gauge Month
200
Rai Evaluation at fine time space resolution using dense gauge
network and MPE radar based product in Louisiana US100 network and MPE radar-based product in Louisiana, US100
Analysis period: (Aug 2004 Dec 2006) The 4x4 km2 MPE and the800 1 2 3 4 5 6 7 8 9 10 11 12
0
M th
Analysis period: (Aug 2004 – Dec. 2006). The 4x4 km2 MPE and the gauge data were re-sampled to the 8x8 km2 CMORPH resolution 600 17 16 %
21.83 %
A l i f Bi
Month
gauge data were re-sampled to the 8x8 km CMORPH resolution
Diurnal cycle along a latitudinal cross section (15o N) 400
17.16 %Analysis of Bias
Diurnal cycle along a latitudinal cross section (15o N)200
400
mm
)
6 24 %200
pth
(m
3.09 % 1.97 %
6.24 %
0.04 %
Based on historical Gauge (Camberlin, 2009)0
l d
ep
1 93 %0 69 % 1 93 %g ( , )
-200
infa
ll -1.93 %-3.09 %-0.69 % -1.93 %
P d d i CMORPHRain gauges -400R
a Hit Bias
Missed RainProduced using CMORPH
-600
Missed Rain
False Rain
T t l Bi
-800
Total Bias-21.43 %-21.10%
CMORPH_Rs MPE_Rs CMORPH_MPE-800
100 4
80
90
3 D i70
80
%)
3
%)
Detection Analysis
60
70
OD
(%
2FD
(% Analysis
50PO
PO
F
Produced using TRMM-3B42 30
40 1
P b bili f dProduced using TRMM 3B42 30
010-1
e
Probability of exceedance0 1 2 3 0 1 2 3
dan
ce
25 25epth
CMORPH on Rs
C O
ceed
20
25
epth
ai
n)
20
25
in D
e
CMORPH on MPE
MPE on Rsof
ex
15
20
ain
De
ota
l ra
15
20
ed R
ara
in)
ilit
y o
10
15
ed R
ao
f to
10
15
etec
teo
tal
r
10-2
ob
abi
CMORPH
Diurnal cycle of rainfall in Lake Tana basin source of 5
10
Mis
se (
%
5
10
ly D
e%
of
to10
Pro
MPE
Diurnal cycle of rainfall in Lake Tana basin, source of 0
5M
0
5
Fal
sel
(%Rs
Upper Blue Nile 0 1 2 30
Th h ld ( h-1)
0 1 2 30
1
F 4 8 12 16 20
1
Upper Blue Nile Threshold (mm h-1) Threshold (mm h-1)Rainfall threshold (mm h-1)
Analysis ConclusionsAnalysis Period: June
Conclusions2
2
CMORPH Period: June-Aug 2007
Both products capture the overall patterns of rainfall climatology over the Nile Basin. Over basin areas north of the CMORPH
TRMM-3B421.5)
TRMM-3B42Rain Gauge Aug, 2007 equator, monthly rainfall amounts are mostly overestimated by CMORPH and underestimated by TRMM-3B42. The 1.5
hr-1
) Rain Gauge
m h
r-1) g
agreement is better around the equator. 1(m
m h
1
all
(mm
CMORPH performed better than TRMM-3B42 in terms of capturing diurnal cycle of rain rate over Lake Tana basin (source 1
infa
ll (
0 5Rai
nfa
p p g y (of Blue Nile). The major limitation is over the mountains and the Lake shore. TRMM-3B42 extremely underestimated rain 0.5R
a0.5
) j yrates in this basin.
00 3 6 9 12 15 18 21 240
Over Lake Victoria basin both products successfully identified regions that have morning rain rate maxima and those that0 3 6 9 12 15 18 21 24
0
Local time (h)
Local time (h)
Over Lake Victoria basin, both products successfully identified regions that have morning rain rate maxima and those that receive their maximum in late-evening to mid-night local times However products significantly underestimated the rain2
2 receive their maximum in late evening to mid night local times. However, products significantly underestimated the rain rate at the middle of the lake.
CMORPH
TRMM-3B42
2
CMORPH
TRMM 3B42rate at the middle of the lake.
CMORPH d ti l IR l d t lt d i l ti l f f th NFS i f ll ff d l
1.5
hr-1
) Rain Gauge1.5
1 )
TRMM-3B42Rain Gauge
CMORPH and an operational IR-only product resulted in relatively poor performance of the NFS rainfall-runoff model. Significant improvements in model performance were gained by adjusting satellite rainfall inputs using gauge data In Blue
1(mm
h
mm
hr-
Significant improvements in model performance were gained by adjusting satellite-rainfall inputs using gauge data. In Blue Nile basin the relative volume error of simulated flow decreased from 40% when using CMORPH instead of IR estimatesai
nfa
ll
1
fall
(m
Nile basin, the relative volume error of simulated flow decreased from -40% when using CMORPH instead of IR-estimates as a model input However merging either product with real-time gauge data resulted in nearly equivalent performance
0.5Ra
0.5Rai
nf
as a model input. However, merging either product with real-time gauge data resulted in nearly equivalent performance. 0
•Evaluation of CMORPH at fine space-time scales showed that: CMORPH has very negligible total bias (0.04%) but its i d i d f l i bi i ifi tl l ( 21 10 d 21 83 %) CMORPH h i f ll d t ti
0 3 6 9 12 15 18 21 24Local time (h)
0 3 6 9 12 15 18 21 240
L l ti (h)
missed-rain and false-rain biases are significantly large (-21.10 and 21.83 %). CMORPH has poor rainfall detection bilit i th t d ti l l f li ht i t (< 2 h 1) C l ti b t CMORPH ti t d
2
CMORPH
Local time (h)
capability in the study area particularly for light rain rates (< 2 mm h-1). Correlation between CMORPH estimates and gauge data increased from 0 4 at hourly scale to 0 6 at a daily scale Similarly the relative RMSE decreased from 840 % of mean1.5)
CMORPH
TRMM-3B42Rain Gauge2 data increased from 0.4 at hourly scale to 0.6 at a daily scale. Similarly, the relative RMSE decreased from 840 % of mean
rain rate at hourly scale to ~200% at 2 days time scale Using radar based MPE as reference results in favorable accuracym h
r-1
Rain Gauge2 CMORPH
TRMM-3B42 rain rate at hourly scale to ~200% at 2-days time scale. Using radar based MPE as reference results in favorable accuracy of CMORPH than its actual accuracy in terms of POFD and falsely detected rain depth It results in less favorable accuracy
1
all
(mm
1.5
r-1) Rain Gauge
of CMORPH than its actual accuracy in terms of POFD and falsely detected rain depth. It results in less favorable accuracy in terms of total bias and hit bias0 5R
ain
fa
1mm
hr
in terms of total bias and hit bias
A k l d t0.51
nfa
ll (
m
Acknowledgment0 3 6 9 12 15 18 21 240 0.5R
ai gFunding provided by the National Science Foundation (NSF) through US Egypt Cooperative Research
Local time (h)
0 Funding provided by the National Science Foundation (NSF) through US-Egypt Cooperative Research Program (Award Number 0914618)
0 3 6 9 12 15 18 21 240
Local time (h)
Program (Award Number 0914618)oca t e ( )