Saving water in farming : methodology for water conservation verification efforts in the agricultural sector

This dissertation develops, tests and validates statistical methods for verifying the amount of water conserved as a result of investments in precision leveling, other on-farm conservation measures in place, weather variation and farmer behavior. This evaluation uses a sample of 328 unique fields from Lakeside Irrigation Division in Texas over a six-year period, totaling 966 observations. Results show that precision leveling accounts for a 0.30 acre-feet reduction of irrigation water per acre leveled. This Mixed-Level Model (MLM) estimate for precision leveling water savings is more precise than the estimates either from an Ordinary Least Square Model or a Fixed Effect Model. A meta analysis combines the results from this model with other similar studies. Although the mean estimate of the meta-analysis is similar to the MLM estimate, the meta-analysis further reduces the standard error of the mean precision leveling estimate by 2 percent. A better approximation of the acre-feet water savings per acre farmed translates into less uncertainty for water regulators, managers and policymakers regarding the volume of conserved water that is available for transfer.