Drought Forecasting for Small to Mid-sized Communities of the Southeast United States
Type of Degreedissertation
MetadataShow full item record
Most of the climate variability in the Southeast United States has been attributed to El Niño Southern Oscillation (ENSO) and this climate variability has resulted in increased the stress on water resources of the region and drought is one of the most expensive outcomes of this climate variability. Drought is a major concern for small to mid-size communities in the Southeast as it poses a serious risk to the performance of water supply systems of such communities and may cause short term failures. In response, this study was undertaken to study the impact of ENSO on precipitation and streamflows and to develop a Community Water Deficit Index (CWDI) for forecasting drought in small to mid-size communities. The usefulness and value of this drought forecast information for water resource managers was then assessed. Results indicated a significant relationship between ENSO and precipitation and streamflow with dry conditions during winter months being associated with La Niña in the southern climate divisions of Alabama. It was found that this information can provide a basis for the water resource managers in Alabama to incorporate ENSO related climate variability in their decision-making. During a low precipitation and high temperature ENSO phase (La Niña), the loss of soil moisture through evaporation increases the dynamic demand of water due to increase in outdoor water use by the residents (lawn irrigation etc.). System Dynamics modeling software STELLA TM was used to develop a model addressing the relationship between water supply and demand of a community and the CWDI was estimated as ratio of available storage and desired level of water storage in the reservoir of the community. The index was tested in two small to mid-size communities in the region and it demonstrated skill in monitoring and forecasting drought. Impact of climate variability on water demand of the community and how the knowledge of this forecast allows mitigation of negative impacts were studied. A multiple linear regression approach was used to predict per capita water demand based on daily precipitation, daily maximum temperature, and a one-day lag variable to account for temporal persistence in time series of water use. It was found that there is considerable accuracy in predicting water use based on climatic variables (R2 values ranging from 0.62 – 0.84). The model was run using historical data to estimate volumetric and cost savings associated with the use of this drought forecast information and it was found that considerable savings could be made by using CWDI to plan ahead thus minimizing the drought vulnerability of community water systems.