Quantifying the Effect of Irrigation on Nonlocal Aspects of the Atmosphere

Abstract

AbstractIrrigation‐induced effects on weather have been studied for several decades using observations and numerical weather prediction models. However, few have ventured to investigate nonlocal, downstream effects on convective weather produced during the passage of strongly forced synoptic systems, or as we refer to as high‐impact weather. One reason for this discrepancy is the influence of chaos seeding (Ancell et al., 2018, https://doi.org/10.1175/BAMS‐D‐17‐0129.1) from the model physics in areas where precipitation or convection are occurring, causing unrealistic changes in precipitation far downstream in a matter of hours. The effects of chaos seeding have made it difficult to distinguish what downstream weather modifications are irrigation‐induced feedbacks and what are products of chaos seeding itself—a distinction at the center of this work. The authors investigate whether atmospheric perturbations produced by land‐surface feedbacks from local irrigation in the Texas Panhandle can significantly modify downstream atmospheric structure on time and space scales across days and thousands of kilometers using the Advanced Research Weather Research and Forecasting Model Version 3.5.1. Irrigation is represented by perturbing the grid point variable of soil moisture with an ensemble of 75 members per 9 case studies for 96‐hr simulations, with 25 members representing unrealistic perturbations applied in New York accounting for chaos seeding. The authors find irrigation influences downstream and nonlocally elsewhere beyond the modification of chaos seeding, with effects shown in three of nine cases amid precipitation shifts occurring around cyclone, dryline, and cold front tracks with changes of 50 mm or more.