It might seem intuitive to a novice physical sciences student: If the amount of moisture in a region's soil is low, there's bound to be less water evaporation from the soil, fewer cloud formations and, correspondingly, less rain in that region. Conversely, there's probably been a lot of rain in areas with high soil-moisture readings.

Recent research at North Carolina State University now provides some scientific validity to this elementary hypothesis by identifying a link between soil-moisture readings and drought indices across North Carolina.

The paper that details this research - written by Drs. Sethu Raman, professor, and Dev Niyogi, research assistant professor of marine, earth and atmospheric sciences at NC State, and master's degree student Aaron Sims - was published in the April 15 issue of the journal Geophysical Research Letters.

Niyogi, who is also assistant state climatologist at the State Climate Office of North Carolina, says the research showing a link between low soil-moisture readings and drought could have far-reaching implications for forecasters' abilities to predict weather. Soil moisture is just one of the many local forces that produce weather.

"Our data links drought, which is really the lack of soil moisture, to soil-moisture readings taken at three areas across North Carolina," Niyogi said. "Now we hope to use this information to predict how drought might evolve." Understanding the evolution of drought can give water resource planners more critical information when they make crucial conservation decisions, for example.

Researchers compared data from two different drought indices, soil moisture readings and precipitation data from three North Carolina Agricultural Network (AgNet) stations in Fletcher, Clayton and Lewiston, N.C., representing the state's mountain, piedmont and coastal plain regions, respectively.

Looking at values over a one-year period, researchers noticed that one drought index, the Standardized Precipitation Index (SPI), followed the actual precipitation data curves rather closely; the other drought index - the Palmer Drought Severity Index (PDSI) - did not capture extreme variabilities in rainfall very well.

"This means that SPI seems to do a better job of reflecting short-term drought variabilities in North Carolina," Niyogi says.

Niyogi and fellow assistant state climatologist Ryan Boyles say that drought really has no single definition; what drought indices reflect in North Carolina might seem really wet in Arizona, for instance. Moreover, there are six or eight prevalent drought indices used; the problem is matching the proper index with the specific needs of a locality or region.

"This research tested just two indices over a short period of time and shows promise that they can help improve weather forecasting," Niyogi says. "Now we need to do even more monitoring and test other indices."

Niyogi and Boyles say plugging drought index data into models and applying it to longer-range weather forecasting is the next step. Having the ability to predict drought - or its termination - could have lasting effects on the state's agrarian population and policy makers' abilities to manage water resources.

- kulikowski -

Editor's note: A copy of the paper is available by contacting Dr. Dev Niyogi at 919/513-2102 or dev_niyogi@ncsu.edu. An abstract of the paper follows.

"Adopting Drought Indices for Estimating Soil Moisture: A North Carolina Case Study"
Authors: Aaron P. Sims, Dev dutta S. Niyogi, Sethu Raman, North Carolina State University
Published: April 15, 2002, in Geophysical Research Letters

Abstract: Soil moisture availability has a significant impact on environmental processes of different scales. Errors in initializing soil moisture in numerical weather forecasting models tend to cause errors in short-term weather and medium range predictions. We study the use of two drought indices: Palmer Drought Severity Index (PDSI) values and Standardized Precipitation Index (SPI) for estimating soil moisture. SPI and PDSI values are compared for three climate divisions: western mountains, central piedmont, and the coastal plain in North Carolina, USA. Results suggest SPI to be more representative of short-term precipitation and soil moisture and hence a better indicator of soil wetness. A regression equation that uses SPI is proposed to estimate soil moisture.