Irrigation demand in a changing climate: Using disaggregate data to predict future groundwater use

The paper estimates an irrigation water demand function using disaggregate climate and well data over a 33 year time period. Aggregating climate information over long periods, like a year, causes a loss of detail on temporal climatic variation, while aggregating climate information over space causes a loss of detail on spatial variation. This analysis uses disaggregate climate variation at a temporospatial level to determine the effects of climate on groundwater use. Results show that increased heat, measured in cooling degree-days, correlates with increased water use, while increased precipitation correlates with decreased water use. However, the effects are generally magnified for later summer months, and are lower at the beginning and end of the growing season – with a few exceptions. Soil type effects groundwater demand for July in particular, and has a ubiquitous effect at the marginal level. Other economic and physical variables were controlled for in the analysis. Using NOAA climate scenarios, which depict climate under increased carbon dioxide for three time periods in the future, we perform an ex-ante analysis using the coefficients derived from our model to determine future irrigation demand. In both high- and low-emissions scenarios, irrigation demand increases. The water increases derived in the forecast range from 10 – 15% from the average in earlier years, to as much as 27% in later years. Advisor: Karina Schoengol