Global climate model simulations with a realistic representation of irrigation

Sensitivity studies with global and regional climate models have shown that replacing existing land cover with irrigated land can cause significant changes in climate. To provide a more realistic assessment of irrigation's influence, we performed a series of irrigation experiments with the Community Atmosphere Model (CAM). An irrigation fraction was computed for each grid cell based on FAO irrigation maps, and this fraction of agricultural area was then irrigated in the cell throughout an AMIP simulation (1980-1999). This experiment and a control simulation were performed both at 2-degree and half-degree spatial resolutions, with the latter aimed at resolving land use within heterogeneous areas such as California. Results are presented for changes in temperature, precipitation, humidity, runoff, and other variabiles. In addition, we evaluated whether irrigation reduced the bias of climatologies when compared to observed climate over the AMIP time period. Finally, we investigate the factors leading to regional differences in the response of climate to irrigation.