River main stem

A mass balance of the river main stem is used to characterize discharge. The mass balance equation (Cohen et al., 2001; Owen-Joyce et al., 1996) can be described as:

Q_ds = Q_us + Q_rf + P_r + T_r - E_r - S_a - Q_sb (1)

Q_dsflow at the downstream boundary Q_usflow at the upstream boundary Q_rfreturn flow to the river P_rprecipitation to river T_rtributary inflow E_revaporation from river open water surface S_achange in aquifer storage Q_sbflow to sub-areas where, Q_ds is flow at the downstream boundary, Q_us is flow at the upstream boundary, Q_rf is return flow to the river, P_r is precipitation to river, T_r is tributary inflow, E_r is evaporation from river open water surface, S_a is change in aquifer storage, and Q_sb is flow to sub-areas.

Discharges at the upstream boundary (Q_us) are from records of two national hydrometric stations. Return flow to the river (Q_rf) consists of agricultural and municipal drainage. Records of agricultural drainage, where available, were obtained from the Water and Salt Monitoring Project. However the agricultural drainage is mixed with spring water. Therefore these data are considered uncertain. Then the estimated agricultural drainage was used in this study, and the available drainage data were used to calibrate the irrigation model of sub-areas. There were no records of municipal water deliveries available, so the municipal water use is estimated from estimated per capita consumption, population, GDP, etc. The population and GDP data were obtained from the local statistics bureau. Estimations of municipal effluent discharge to the river were based on the assumption that 35% of municipal water use became effluent discharge to the river. It is worth mentioning that the volume of municipal drainage water (8×10⁶ m³/year) is small in comparison with the volume of water discharged at the upstream boundary (8,000×10⁶ m³/year).

Precipitation to rivers (P_r) is calculated from averaged precipitation records of the three hydrometric stations and open water surface area from the CBERS-1 images. The tributary runoff (T_r) is usually generated from mountainous area outside the plain, and flows into the study area as mountain torrent and spring. There are not gauged precipitation data for the runoff generating area, thus the results of the mountain torrent and spring ravines investigation in 1999 were used as the basis of the tributary runoff. The Akesu Hydrological Bureau (AHYB) estimated runoff by extending the runoff isopleth map to the mountainous area and correlating this with an extended isohyetal map to project an annual tributary runoff of 600×10⁶ m³. The projection is roughly consistent with the result of the investigation in 1999 (560×10⁶ m³). Therefore, the spatial and temporal distributions of annual runoff were taken from the 1999 investigation and were scaled to fit the projection of AHYB (600×10⁶ m³/year). Evaporation from river open water surface (E_r) is calculated from averaged pan evaporation records of the three hydrometric stations and open water surface area from the CBERS-1 images. Estimation of change in storage in the alluvial aquifer ( S_a) was based on an assumption that the storage change is in direct proportion to the averaged stream flow as follow:

S_a = (2)

regression coefficient where, is the proportionality constant derived from linear regression analysis of available observations. The storage change in the alluvial aquifer is connected to the groundwater of two sub-areas, meaning that the river-groundwater interaction could affect the groundwater (i.e. water table depth) of the alluvial plain. Water from the river main stem (Q_sb) has discharged into the Tuoshigan-Kumalake River plain as water diversion for irrigation. The net water diversion data were available from ARAA. [5]GDPGross Domestic Product [6]AHYBAkesu Hydrological Bureau