* Click image to expand

Linked NASA-NOAA Application of NARR-Based NLDAS Ensemble Simulations to Continental-Scale Drought Monitoring

Investigators: Brian A. Cosgrove (PI, NASA/GSFC/SAIC) and Charles Alonge (NASA/GSFC/SAIC)

Funded by the NOAA Climate and Global Change Program (CPPA Element) with assistance from NASA's Water Management Program

Abstract Government estimates indicate that droughts cause billions of dollars of damage to agricultural interests each year. Resource managers act to mitigate the impact of such events, but are limited by the characterization of such droughts: their response can only be as accurate as the characterization of the severity, temporal, and spatial extent of these climatic events. More effective identification of droughts would directly benefit decision makers, and would allow for the more efficient allocation of resources that might mitigate the event. Land data assimilation systems, with their high quality representations of soil moisture, present an ideal platform for drought monitoring, and offer many advantages over traditional modeling systems. The recently released North American Regional Reanalysis (NARR) covers the NLDAS domain and provides all fields necessary to force the NLDAS for 27 years. This presents an ideal opportunity to combine NARR and NLDAS resources into an effective real-e drought monitor. Toward this end, this project seeks to validate and explore the NARR's suitability as a base for drought monitoring applications, both in terms of data set length and accuracy. Along the same lines, the proposed research will examine the impact of the use of different (longer) LDAS model climatologies on drought monitoring, and will explore the advantages of ensemble simulations versus single model simulations in drought monitoring activities. An additional benefit to the modeling community will be the creation of NARR- and observation-based high quality 27 year, 1/8th degree, 3-hourly, land surface and meteorological forcing data sets. An investigation of the best way to force an LDAS-type system will also be made, with traditional NLDAS and NLDASE forcing options explored.

NASA Products

Remote Sensing Inputs: MODIS, AMSRE, TRMM
Model: LIS

Project Partners

NOAA   /   NCEP

Decision Support Tools

The Drought Monitor post-processor will be constructed once forcing data and output is in place, and will use this data to compute several drought indices. While it is important for the drought monitor to leverage its unique ensemble LDAS-based strengths to explore new LDAS-based drought indices, it is equally important that index output be verifiable against established and time-tested drought monitoring efforts. As such, the post-processor will produce both new LDAS-based indices as well as indices commonly used in the operational and research community to characterize hydrological, meteorological, and agricultural drought.

In particular, SPI and PDSI indices will be used as gauges of meteorological drought, while the PHDI and TWD hydrological drought indices will be used to characterize hydrological drought. The Palmer Z index, and the Sheffield et al. (2004) percentile soil moisture index (VIC index) will be used to assess agricultural drought. The proposed 27-year 1/8th degree forcing data set will serve as the basis of the meteorological and Palmer drought indices, while output from the ensemble LDAS LSM simulations will serve as the basis of the TWD and soil moisture percentile measures. Each of these drought characterization methods represents an established means of drought indexing, and either is available from long term archives.

In particular, the Palmer indices are available from NCDC; TWD values will be computed from archived USGS stream flow measurements; SPI archives are available from the University of Nebraska; and archives of the VIC index are available from the University of Washington. Capitalizing on the strengths of LDAS LSM output, two experimental LDAS drought indices will also be explored. The first measure will utilize CLM3's strong vegetation sub-model to produce prognostic LAI values and generate NDVI and Vegetation Condition Index output. This will be directly comparable to NOAA and NASA remotely sensed NDVI-based VCI fields, and will assess agricultural drought. The second experimental index will take the form of an LDAS-based Palmer index. It will utilize the LSM ensemble output and NARR forcing data to calculate weekly and monthly moisture anomalies. Dynamically calculated duration factors from the recently developed Self Calibrating Palmer Index (SCPI, Wells et al. 2004) will then be used to create enhanced Z-index, PDSI, and PHDI index fields, assessing all three types of drought. Experimental Palmer values will be compared to archives of standard Palmer values. These indices will take advantage of the ensemble nature, vertical resolution, complete spatial coverage, and wide range of output fields offered by the LSMs run within LIS.

Reports

top

Date Last Modified: 08/11/09

 + NASA Privacy Statement, Disclaimer,
and Accessibility Certification