Transcript Bias-corrected Satellite Estimates - ISAC

High Resolution Gauge – Satellite Merged
Analyses of Precipitation: A 15-Year Record
Pingping Xie,
Soo-Hyun Yoo, Robert Joyce, Yelena Yarosh, Shaorong Wu, and
Roger Lin
2012.10.16.
Objectives
•
Reprocessing CMORPH for the entire TRMM era
(1998-present) using a consistent version
algorithm and fixed versions of inputs throughout
the data period
•
Performing bias correction / magnitude adjustment
•
Blending bias-corrected CMORPH with gauge
analysis
CMORPH Reprocessing
1) Algorithm and Inputs
•
Algorithm
•
•
•
•
CMORPH algorithm as of 2009
Joyce et al (2004)
No KF enhancements
Inputs
•
•
•
PMW L2 retrievals (GPROF 2004)
CPC Geo sat IR at 4km (Janowiak et al. 1999)
NESDIS daily snow maps
CMORPH Reprocessing
2) Products


8kmx8km / 60oS-60oN
30-min interval / January 1998 to the present
CMORPH Bias Correction
1) Bias in the raw CMORPH
Regionally different
Temporally changing
Non-linear
CMORPH Bias Correction
2) Strategy
•
Over Land
– PDF matching against daily gauge analysis
– PDF tables established as a function of region and season
using historical and real-time data
•
Over Ocean
– Calibration against a long-term record
(pentad GPCP) with stable quality but
coarser resolution (2.5olat/lon, 5-day)
CMORPH Bias Correction
1) Results over land
2000-2009 Annual Mean
Daily Gauge
 Large-scale bias removed
 Correlation improved
KF-CMORPH
Comparison over Africa
Gauge-Adjusted KF-CMORPH
CMORPH Bias Correction
1) Comparison with daily gauge for June 2011
Correlation
Bias
Region
CPC
Original
CPC
HS CRTD
CPC
RT CRTD
CPC
Original
CPC
HS CRTD
CPC
RT CRTD
Globe
0.551
0.617
0.647
0.098
-0.247
-0.171
60N-40N
0.535
0.549
0.587
0.003
-0.481
-0.142
40N-20N
0.578
0.650
0.677
0.848
0.177
-0.135
20N-20S
0.553
0.584
0.602
-0.512
-0.540
-0.261
20S-40S
0.605
0.715
0.767
-1.128
-0.477
-0.256
40S-60S
0.666
0.684
0.698
-2.755
-0.921
-0.467
Bias in mm/day
Bias reduced substantially in CPC version of the estimates
CMORPH Bias Correction
1) Applications in verification of CFSR precipitation
Combining Bias-Crtd CMORPH with Gauge
1) Strategy
–
This is only possible for several regions due to
different daily ending time in the gauge reports
•
•
•
•
•
–
Africa
CONUS/MEX
S. America
Australia
China
(06Z)
(12Z)
(12Z)
(00Z)
(00Z)
Based on Xie and Xiong (2011)
Combining the bias-corrected CMORPH with gauge
observations through the Optimal Interpolation
(OI) over selected regions where gauge
observations have the same daily ending time
•
CMORPH and gauge data are used as the first guess and
observations, respectively
Combining Bias-Crtd CMORPH with Gauge
2. Example
•
Gauge analysis depict
heavy rain but tend to
extend the raining
area
•
Satellite data tend to
under-estimate
•
Merged analysis
present improved
depiction of the heavy
rain
Combining Bias-Crtd CMORPH with Gauge
3. Validations
 Reports from up to 28 stns available in
a 0.25olat/lon grid box over Seoul

Arithmetic mean of 28-stn reports
taken as the ‘truth’

Gauge analyses using 1000
combinations of sub-set stations
are combined with bias-corrected
CMORPH and compared to the
‘truth’ at the grid box over Seoul
 Correlation is calculated for the
combined analyses and the input
gauge / CMORPH. Results are
plotted in different colors for
different gauge network densities
Summary
• Three sets of gauge-satellite precipitation analyses
• Reprocessed CMORPH Satellite Estimates
•
Global
•
8kmx8km; 30-min
•
1998 to the present
• Bias-corrected Satellite Estimates
•
Global
•
8kmx8km; 30-min
•
1998 to the present
• Gauge-satellite combined analyses
•
Regional
•
0.25olat/lon; daily
•
1998 to the present