Simulating the Impacts and Sensitivity of the Southeastern United States Climatology to Irrigation

The diurnal variations from a high-resolution regional climate model (Regional Spectral Model; RSM) are analyzed from 6 independent decade long integrations using lateral boundary forcing data separately from the National Centers for Environmental Prediction Reanalysis 2 (NCEPR2), and European Center for Medium-Range Weather Forecasts (ECMWF) 40-year Reanalysis (ERA40) and the 20th Century Reanalysis (20CR). With each of these lateral boundary forcing data, the RSM is integrated separately using two convection schemes: the Relaxed Arakawa-Schubert (RAS) and Kain-Fritsch (KF) schemes. The results show that RSM integrations forced with 20CR have the least fidelity in depicting the seasonal cycle and diurnal variability of precipitation and surface temperature over the Southeastern United States (SEUS). The remaining four model simulations show comparable skills. The differences in the diurnal amplitude of rainfall during the summer months of the 20CR forced integration from the corresponding NCEPR2 forced integration, for example, is found to be largely from the transient component of the moisture flux convergence. The root mean square error (RMSE) of the seasonal cycle of precipitation and surface temperature of the other four simulations (not forced by 20CR) were comparable to each other and highest in the summer months. But the RMSE of the diurnal amplitude of precipitation and the timing of its diurnal zenith were largest during winter months and least during summer and fall months in the four model simulations (not forced by 20CR). The diurnal amplitude of surface temperature in comparison showed far less fidelity in all models. The phase of the diurnal maximum of surface temperature however showed significantly better validation with corresponding observations in all of the 6 model simulations The impacts of irrigation on SEUS diurnal climate are then investigated. An extreme case is assumed, wherein irrigation is set to 100% of field capacity over the growing season of May through October (IRR100). Irrigation is applied to the root zone layers of 10-40cm and 40-100cm soil layers only. It is found that in this regime there is a pronounced decrease in monthly averaged temperatures in irrigated regions across all months. In non-irrigated areas a slight warming is simulated. Diurnal maximum temperatures in irrigated areas warm, while diurnal minimum temperatures cool. The daytime warming is attributed to an increase in shortwave flux at the surface owing to diminished low cloud cover. Nighttime cooling results as a consequence of higher net downward ground heat flux. Both diurnal and monthly average precipitations are reduced over irrigated areas at a magnitude and spatial pattern similar to one another. Due to the excess moisture availability, evaporation is seen to increase, but this is balanced by a corresponding reduction in sensible heat flux. Concomitant with additional moisture availability is an increase in both transient and stationary moisture flux convergences. However, despite the increase, there is a large-scale stabilization of the atmosphere stemming from a cold surface and a warmed vertical column. Three additional regional climate model runs centered on the SEUS assume a crop growing season of May through October and are irrigated at 25%, 50%, 75% (IRR25, IRR50, IRR75, respectively) of the root zone field capacity to assess the sensitivity of the SEUS climate to irrigation. A fifth run, assuming no irrigation (CTL), is used as the basis for comparison. Across all IRR runs, it is found that there is a general reduction in monthly mean precipitation over the irrigated cells relative to CTL, with much of the change occurring in the sub-diurnal scales. This manifests as an increase dry days and reduction in > 1 mm/day rainfall events. IRR25 is seen to have the lowest change in both, while IRR100 is seen to have the greatest change. Area-averaged precipitation over the irrigated cells reveals a strong reduction in precipitation in IRR100 (on the order of 0.4 mm/hr) with a much weaker reduction in IRR25. Vertically integrated moisture convergence is seen to have the most pronounced sensitivity pattern across all runs. Monthly averaged temperatures are reduced over irrigated areas, with the intensity of the reduction increasing as irrigation vigor increases. This is attributed to a systematic change in ground heat flux that transports heat into the subsurface soil layers in the irrigated cells. The precipitation ahead of the transient cold fronts is reduced by irrigation as it passes over the irrigated cells. The intensity of the net precipitation reduction becomes more intense as irrigation vigor increases. Lastly, heat waves in the SEUS are reduced in intensity just over the irrigated cells, though likely increasing in frequency due to lowered temperature thresholds for heat wave definition.