1. Towards a Near Real Time Drought monitoring based on NCEP Regional Reanalysis Muthuvel Chelliah, Kingtse Mo and Wayne Higgins Climate Prediction Center, NCEP/NWS/NOAA An Experimental Drought Early Warning System (DEWS ) An integrated component of national policy to monitor and predict drought in support of the NIDIS An integrated component of national policy to monitor and predict drought in support of the NIDIS An early drought warning system will mitigate the impact of drought over the United States and the improved operational drought monitoring will contribute to the National Integrated Information System (NIDIS); 1. An early drought warning system will mitigate the impact of drought over the United States and the improved operational drought monitoring will contribute to the National Integrated Information System (NIDIS); 2. With the recent accomplished Regional Reanalysis (RR) and the North American Land Data Assimilation System (NLDAS), we are able to build a consistent mesoscale drought monitoring system; 3.With the NCEP CFS (Coupled Forecast System) forecasts, it is possible to develop and implement an experimental DEWS based on the dynamical forecasts and regional analysis/NLDAS in a near future. Monitoring Drought More than one index or variable is needed to give a complete picture A) INDICES: (i) Standard Precipitation Index (i) Standard Precipitation Index Based on precipitation (P) alone; Based on precipitation (P) alone; Easy to extend to forecasts; Easy to extend to forecasts; Does not include soil/hydrological conditions. Does not include soil/hydrological conditions. (ii) Palmer Drought Severity Index (PDSI) (ii) Palmer Drought Severity Index (PDSI) Based on the water balance equation; Based on the water balance equation; Difficult to extend to forecasts; Difficult to extend to forecasts; B) CIRCULATION & SURFACE FIELDS Both the SSTs (remote) and soil moisture (regional) can Both the SSTs (remote) and soil moisture (regional) can influence drought. influence drought. (i) SSTs can be obtained from the CDAS2. (i) SSTs can be obtained from the CDAS2. (ii) The following fields and anomalies are based on (ii) The following fields and anomalies are based on the RR. E, PE, soil moisture & soil temperature at 4 the RR. E, PE, soil moisture & soil temperature at 4 layers, solar radiation, long wave radiation at the layers, solar radiation, long wave radiation at the surface, sensible heat, T2m, moisture fluxes and surface, sensible heat, T2m, moisture fluxes and divergence, runoff, and snow conditions. divergence, runoff, and snow conditions. PDSI from the RR The same basic frame work as Palmer (1965) is adopted: (i) The water balance equation; (i) The water balance equation; (ii) The difference between P and the expected P from the (ii) The difference between P and the expected P from the maximum conditions (CAFEC); maximum conditions (CAFEC); (iii) The assumption of the first order Markov process; (iii) The assumption of the first order Markov process; But, the following changes are made: (i) Monthly mean PE, E, runoff, total soil moisture change (i) Monthly mean PE, E, runoff, total soil moisture change were taken directly from RR ; were taken directly from RR ; Potential Recharge is defined as PR=Smax-S ; where Potential Recharge is defined as PR=Smax-S ; where Smax is the maximum total soil moisture for a given Smax is the maximum total soil moisture for a given calendar month; S is the total soil moisture at the calendar month; S is the total soil moisture at the beginning of the month; Potential precipitation is beginning of the month; Potential precipitation is assumed to be the maximum P for a given calendar assumed to be the maximum P for a given calendar month; month; (ii) The AWC and assumption of two soil layers are no (ii) The AWC and assumption of two soil layers are no longer needed. Normalization is recalibrated. longer needed. Normalization is recalibrated. PE from Palmer and RR Improvement for some fields are substantial: e.g. PE (Thornthwaite) does not capture the west-east contrast and underestimates PE (RR). Palmer PDSI (RR), PDSI (Palmer) and P (anom) averaged over large areas By and large, the PDSI(RR) averaged over a large area is fairly close to the PDSI based on original Palmer (1965); The PDSI(RR) compares well with P-anom and weights soil moisture more heavily. Large differences occur over the western mountain region Composites of PDSI based on SPI6 index For the PNW the PDSI(RR) shows detailed orographical features that the PDSI(Palmer) from the Climate Division Data misses. For the Southern Plains, the PDSI(Palmer) only indicates mild drought (below -1) while the PDSI(RR) is below -2 (moderate and severe drought). RR Composites of PDSI (Palmer), PDSI(RR) and normalized P anom were formed when the SPI6 for the Paciifc NorthWest (PNW) and S-Plains are below -1 (indicating moderate and severe drought). PDSI (RR) By and large, the PDSI (RR) averaged over a large area or over a long period is characteristically similar to the PDSI (Palmer) based on the climate division data. However, the advantages of PDSI (RR) are: A) It can resolve mesoscale features because RR has a A) It can resolve mesoscale features because RR has a horizontal resolution of 32km; horizontal resolution of 32km; B) Gives more weight to soil moisture anomalies; B) Gives more weight to soil moisture anomalies; C) More representative and consistent with P and other C) More representative and consistent with P and other fields because they are all taken from the same fields because they are all taken from the same Regional Reanalysis. Regional Reanalysis. D) As the RR’s resolution gets better (than 32km) with D) As the RR’s resolution gets better (than 32km) with time, so does the PDSI (RR). time, so does the PDSI (RR). Monitoring Drought SPI (3-month, 6-month, 12- month) from the Unified SPI (3-month, 6-month, 12- month) from the Unified Gauge Precipitation analysis from 1949-present; Gauge Precipitation analysis from 1949-present; PDSI based on the RR, and NLDAS; PDSI based on the RR, and NLDAS; Total soil moisture and soil moisture from 0-10 cm and Total soil moisture and soil moisture from 0-10 cm and 10-200cm, soil moisture changes; 10-200cm, soil moisture changes; Atmospheric hydrological cycle: E, P, D(Q), PE, runoff, Atmospheric hydrological cycle: E, P, D(Q), PE, runoff, and surface fluxes & radiation terms. and surface fluxes & radiation terms. Fields from the NLDAS will be added Fields from the NLDAS will be added Drought: 2000-2002 Even as early as July 1999, the United States experienced an intensifying drought and heat wave, and affected about 88 counties in six states (KY, MD, OH, PA, VA, and WV - NCDC, 13 Aug 1999). As of early Aug. 1999, the governors of three states (MD, PA and NJ) had declared drought emergencies. Even though September hurricanes Dennis and Floyd ended the short term drought where the excessive rains fell, the extreme long term drought conditions continued over a large area and lingered well into 2002. The following slides illustrate how the various indices capture the many aspects of this drought. Even as early as July 1999, the United States experienced an intensifying drought and heat wave, and affected about 88 counties in six states (KY, MD, OH, PA, VA, and WV - NCDC, 13 Aug 1999). As of early Aug. 1999, the governors of three states (MD, PA and NJ) had declared drought emergencies. Even though September hurricanes Dennis and Floyd ended the short term drought where the excessive rains fell, the extreme long term drought conditions continued over a large area and lingered well into 2002. The following slides illustrate how the various indices capture the many aspects of this drought. Drought monitoring based on the SPI6 (SPI < -1 indicates drought) Drought monitoring based on PDSI (RR). [PDSI (RR) less than -2 indicates drought ] Different indicators for the water year 2002 PDSI (RR)PDSI (Palmer) 1) PDSI(RR) and PDSI(Palmer) both indicate drought over the Southwest and the East, but the PDSI(RR) indicates more severe drought over the East because it weights soil moisture more. 2) PDSI(RR) does not indicate drought over the Northwest consistent with positive P anomalies. 3) SPI6 is similar to normalized P index. Summary and Future Work We are in the process of developing a We are in the process of developing a dynamically-based DEWS based on the dynamically-based DEWS based on the mesoscale Regional Reanalysis and NLDAS. mesoscale Regional Reanalysis and NLDAS. In addition to precipitation, soil moisture from In addition to precipitation, soil moisture from 0-200cm and four layers, we plan to use the 0-200cm and four layers, we plan to use the PDSI based on RR, SPIs and surface fluxes PDSI based on RR, SPIs and surface fluxes and energetics to monitor and forecast and energetics to monitor and forecast drought in near real time. drought in near real time. Objectivity and scalability of RR based drought Objectivity and scalability of RR based drought monitoring. monitoring. The advantages are (a) the RR and NLDAS are The advantages are (a) the RR and NLDAS are mesoscale, (b) all fields are consistent. mesoscale, (b) all fields are consistent. Since drought means various things to various Since drought means various things to various things to various people, more than one things to various people, more than one index is needed to monitor drought index is needed to monitor drought