DROUGHT ACTIVITIES AT NOAAS SOUTHERN REGIONAL CLIMATE CENTER

1
DROUGHT ACTIVITIES AT NOAAS SOUTHERN REGIONAL
CLIMATE CENTER
Luigi Romolo, NOAAs SRCC Regional Climatologist
2

• WHAT CAN THE SRCC
• DO TO IMPROVE THE
• UNDERSTANDING OF
• DROUGHT ???
• Drought Research
• Product Development

3

• Outline for today
• Drought Research
• Internal LSU Faculty Research Grant
• Foster a research partnership with NIDIS
• Identifying synoptic controls on drought
• Identifying the broad-scale atmospheric controls
• TRACS Project (pending funding)
• ENSO-based streamflow forecasting tool
• Developed by the LMRFC
• Aim to transition the application to an
  operational climate service setting

4
Drought Research
Synoptic Controls on Drought Indices in the
Southeastern United States Present and Future
5
Study Area 20- 40N 110- 75W
6
Synoptic window
Synoptic level
Day 1
Day n
TIME SERIES
7
F(x)
Surface Environment
8

• Surface Environment
• Temperature Precipitation
• Degree-Days
• Palmer Drought Severity Index
• Keetch-Byram Drought Index
• Standardized Precipitation Index
• Visual Greenness Index

9
Frequency Distribution example
SAMPLE
10
Example for Precipitation 16 Types 10 dry, 6 wet
SAMPLE
11
Current
Future (GCMs)
Scale Issue
NCEP/NCAR
CGCM
Future Teleconnections Same ? or Different ?
Teleconnections
?
12
Potential for Product Development

• Results can be tied into a Numerical Weather
  Prediction Model or Land-Atmosphere Transfer
  Scheme
• Results can be tied into a Neural Network
• By understanding the broad-scale controls on
  drought we increase our ability to forecast
  drought

13
Drought Product Development
Enhanced Seasonal Forecast of Hydrological
Conditions for Drought Planning and Mitigation
14
TRACS PROPOSAL Joint Project between the SRCC
LMRFC
SRCC
LMRFC
15

• Developed by
• LMRFC originally
• for the Pearl Basin
• Manual in nature
• Plan to automate it and develop a web-based
  interface
• Expand it to a region and then national scale
  (2700 stations)

16

• Forecast /Products
• historical composite tables
• seasonal streamflow
• monthly streamflow
• 75 (25) percent exceedence (non-exceedence)
• lead times 0.5 to 12.5 months
• Methodology uses a NOAA/NWS developed composite
  analysis (Marina Timofeyeva)
• Data USGS streamflow values, CPC Nino 3.4 SST
  probabilities

17
Seasonal Forecasting of Streamflow
This produces a composite table similar to
Historical Composites are produced by dividing
the Below, Near, and Above normal values by the
total number of ENSO years
18
Historical Composites Table
19
Seasonal Forecasting of Streamflow
Historical composite values are combined with the
CPC Nino 3.4 SST probabilities to produce
forecasts for streamflow (0.5 to 12.5 month lead
time) P(BELOW) P(ENB)P(CPCA)P(NB)P(CPCN)P(LN
B)P(CPCB) P(NORMAL) P(ENN)P(CPCA)P(NN)P(CPCN
)P(LNN)P(CPCB) P(ABOVE( P(ENA)P(CPCA)P(NA)P
(CPCN)P(LNA)P(CPCB)
20
Probabilities at each site are tested for
statistical significance using a hypergeometric
function 90 confidence level is used If any of
the 9 episode-category combinations are
significant, then forecasts can be
issued Otherwise an Equal Chances forecasted
is issued
21
INCREASED SKILL
INCREASED SKILL
DECREASEDSKILL
22
(No Transcript)
23
SRCC Computing Power BEOWOLF CLUSTER
41 nodes   Benchmark 0.1725
TeraflopsRanking In 2002, SuperMike was ranked
as the 11th fastestsupercomputer in the world.
Among academic institutionsworldwide, SuperMike
was ranked 2nd.