Material and methodology

The study developed a water balance for the Tuoshigan-Kumalake River plain using existing flow records, estimated evaporation, calculated agricultural drainage, and estimated outflows. Discharge records for the main stem at the Xiehela and Shaliguilanke station were supplemented by local gauge station records for the Xidaqiao station and agricultural diversion records obtained from the Akesu River Administrative Agency (ARAA). Loss due to evaporation in the conjunction plain was calculated by the runoff-evaporation (RE) model described below. Atmospheric forcing data such as temperature, sunshine duration, and wind speed were collected from the Wushi and Wensu stations, two meteorological stations within the study area. Agricultural diversion records from 1980 to 1998 were available. Even though these data are not fully reliable, they were still used to provide the initial conditions for the RE model in this study. The agricultural diversion records after 1998 were provided by a Water and Salt Monitoring Project and were deemed reliable, limiting the study to the period 1999-2002. [4]ARAAAkesu River Administrative Agency

The study area was divided into three hydrological sub-systems: the Tuoshigan-Kumalake River main stem, the Wushi sub-area (sparse dotted area in Fig. 2) and the Wensu sub-area (dense dotted area in Fig. 2). There are surface and ground water connections among these three hydrological sub-systems. Water balance is calculated for every month during the study period.