Optimal reservoir operation and risk analysis of agriculture water supply considering encounter uncertainty of precipitation in irrigation area and runoff from upstream

Abstract

The precipitation in irrigation area and the runoff from upstream can reflect the natural and irrigated agriculture water supply situations. The wetness–dryness encounter uncertainty of precipitation and runoff, affected by the different microclimate in the basin, can lead to the potential water supply risk of reservoir dispatching, which poses a threat to food security. This study develops an innovative integrated modeling framework by coupling Copula joint distribution of precipitation and runoff, the reservoir multi–objective optimal operation model, and multi–criteria decision–making CECG–VIKOR method to assess agriculture water supply risk in the Baojixia Irrigation Area (BIA) of Northwest China. The results indicated that the integrated methodology is an important tool for assessing agriculture water supply risk, which has comprehensive, advanced, and efficient features. The correlation between the precipitation in BIA and runoff from upstream is weak, which can lead to a high uncertainty of the natural and irrigated agriculture water supply conditions. The Clayton copula was more suitable for estimating the joint distribution of precipitation and runoff. And there was a large probability of an unfavorable water balance situation in the BIA, with the extreme water shortage encounter condition of precipitation and runoff had the maximum joint distribution probability of 23.50%. The 9 scheduling schemes were set up according to encounter scenarios and whose optimal scheduling results are R104, R116, R66, R95, R5, R28, R117, R89, and R110, respectively. Among the 9 optimal scheduling results, the water supply risk in BIA increased with the decrease of available water supply and the increase of water demand. Below middle and middle water supply risk were mainly faced by scheme A and B with an occurrence probability of 62.47%, while scheme C had a high risk with an occurrence probability of 37.53% in BIA. The research results can provide a scientific basis for reservoir dispatching risk research and new ideas for making dispatching plans.