Drought Monitoring: progress and challenges Kingtse Mo and

1
Drought Monitoring progress and challenges

• Kingtse Mo and Partners
• Climate Prediction Center
• NCEP/NWS/NOAA

2
Outline

• Every month, CPC issues monthly and seasonal
  drought outlook and participates in the Drought
  Monitor operation over the United States
• To support operational functions, we monitor
  hydroclimate conditions and give drought briefing
  each month to review the current drought
  conditions and drought forecasts
• Satellite data were used to improve Precip and
  NLDAS
• Challenges to cover the global drought

3
Current Partners

• CPC Kingtse Mo, LiChuan Chen, Muthuvel Chelliah,
  Wesley Ebisuzaki
• EMC NLDAS Team Youlong Xia, Jesse Meng, Helin
  Wei, Michael Ek
• NASA/GSFC Randy Koster, Greg Walker
• Princeton Univ. Eric Wood, Justin Scheffield
• Univ. of Washington Dennis Lettenmaier, S.
  Shukla, Francisco Munoz-Arriola
• Web Masters Joe Harrison
• RFCs James noel, Kevin Werner, Andy Wood, SERFC
• Project Funded by NOAA CPPA, TRACS NASA

4
Drought Indices

• More than one index to monitor drought
• Meteorological drought Precipitation deficit.
  (SPI index)
• Hydrological drought Streamflow or runoff
  deficit (SRI index)
• Agricultural drought Total soil water storage
  deficit or soil moisture at the root zone deficit
• (Total soil moisture percentile)
• Runoff and soil moisture Limited data
  available so we need GLDAS

5
SPI

• SPI3 shows short term drought dryness over the
  Great Plains, Southeast
• Wetness over the Midwest
• For longer term SPIs
• Dryness persisted for 6-months or longer
• Typical ENSO signal with dryness over the
  Southern States and Wetness over the North

Warning from the NWS Midwest has been wet for
more than 6 months and possible for Spring floods
D3 D2 D1
6
SPI and other indices

• SPI Advantages
• Easy to use and only need station data
• Cover all time scales
• Do not need a hydrological model. (Other indices
  are model derived products)
• Can cover global (NCDC uses station data and CPC
  uses gridded data)
• SPI Disadvantages
• Not contain snow information
• Areas where soil moisture feedback is important
  or large E, SPI may not be representative (e. g.
  Amazon)

7
All three indices do pick up the major drought
events
SRI3
SM percentiles

Dry Southern states Wet Midwest and Northeast
and the west coast

Drought Indices should be able to pick up major
drought events
8
Uncertainties in the NLDAS impact on regional
applications
U Washington
Ensemble SM

EMC

• The patterns are similar, but there are
  differences
• Over Southeast, the UW does not show anomalies,
  but the EMC does
• Over AZNM, drought depicted by the UW is stronger

9
The EMC NCEP system

• Four models Noah, VIC, Mosaic and SAC
• Climatology 1979-2007
• On 0.125 degrees grid
• P forcing From the CPC P analysis based on rain
  gauges with the PRISM correction.
• Other atmospheric forcing From the NARR

The University of Washington system

• Four models Noah, VIC, SAC and CLM
• Climatology 1915-2007
• On 0.5 degrees grid
• P, Tsurf and low level winds from NOAA/NCDC
  co-op stations
• P from index stations

10
Sensitivity to Precip data

• The RMS difference (Fig.d) between the ncep and
  the UW ensemble SM are large over the western
  U. S. (gt 20).
• Largest differences occur after 2001 as
  indicated by the mean differences for two periods
  (Fig. f and g)

11
Number of station reports averaged over a year
12
CPC Gauge-Satellite Merged Precip Analysis
Xie et al (2011)

• Gauge-based analysis
• OI of reports at 30K stations
• CONUS 0.125olat/lon from 1948
• Global 0.5olat/lon from 1979
• Poor quality over gauge sparse regions
• CMORPH Satellite Estimates
• Integration of all available satellite IR and
  microwave observations
• 8kmx8km over global land (60oS-60oN)
• 30-min time resolution from Jan. 98
• Bias and random error
• Gauge-satellite merged analysis (available around
  summer)
• Bias-corrected CMORPH through PDF matching
  against gauge data
• Same time / space resolution / coverage as CMORPH
• Gauge-CMORPH combined analysis
• Daily analysis / 0.25olat/lon

13
To develop global DEWS, we need the following

• Better P insitu data and better real time
• reporting
• 2 Satellite derived P or radar data but need
  better QC and better calibration. (e. g.
  CMORPH/RMORPH)
• 3.Downward radiation,
• 4.soil moisture data for verification
• 5. Better snow information
• Satellite derived E and SM can help

14
What do we need from GLDAS?

• Better winter time snow properties SWE and snow
  melt

15
Short wave radiation used for the NLDAS forcing
was corrected Better DSWRFgt better Egt better
partition between E and runoff
RUNOFF
NARR
NLDAS
16
Conclusions

• Requirement for drought indices They should be
  able to select all major events.
• For runoff and soil moisture, there are few data
  sets available, we need to use the Global Land
  Data Assimilation System (GLDAS)
• To have better GLDAS products, we need to have
  better Precipitation, downward solar radiation, E
  .
• We also need soil moisture measurements to
  validate the GLDAS products.

17
Drought forecasts

• SM and runoff from lead 1-3 months fcsts
• A) U Washington ESP VIC nested in the CFS
  monthly fcsts with ESP
• B) EMC/Princeton Bayesian corrected Precip from
  the CFS monthly mean fcsts to drive VIC
• C) OHD- SAC model driven by CFS monthly mean
  fcsts to produce streamflow fcsts
• D) CPC- BCSD corrected SPI and SM fcsts from the
  CFS v2
• E) NSIPP model soil moisture outputs

18
What do we need?

• Some thing NOT based on the CFS forecasts.

SE( 26-37N,77-89W)
High resolution forecasts for regional
operations 1, Better Global forecasts 2.Better
high resolution P analyses so we will have better
initial conditions 3.. Better observations for
calibration 4. downscaling 5. Will ensemble
forecasts help? If so, what is the best way to
make ensembles
19
P anom
Dashed line monthly mean anomaly, Solid line-
6-mo running mean

• P has high frequency (HF) and low frequency (LF)
  component.
• LF 6 mo running means
• NCEP P anomalies have large values and
  variances than UW.
• Before 2001, large differences are in HF bands
• After 2002, consistent differences in LF band
• Next LF P? SM changes

20

• Differences between two systems are larger than
  the spread among members of the same system
• The differences are not caused by one model. They
  are caused by forcing.
• In general, extreme values from the UW (Green)
  are larger than from the NCEP (red)

standardized SM anomalies for area
38-42N,110-115W
NCEP(red),UW(green)
21
A dry region
SM has much lower freq. over the western region
22
A wet region
6 mo running mean black line
drought
3 mo running mean (black line)
No smoothing
Red line monthly mean, no smoothing

SM 1-2 months delay
23
Conclusions

• Reliability The spread among the NLDAS driven by
  the same forcing is small. For NLDAS driven by
  different forcing, differences are larger.
  Different systems are able to capture overall
  drought/floods but the severity is uncertain.
• Consistency All different indices derived from
  the NLDAS are able to select strong drought
  events.
• Availability All NLDAS systems are operational
  in near real time.
• What do we need Better real time reporting of
  precipitation from stations and better
  precipitation analyses

24
Number of reports /month averaged over the box
Large drop in real time