The runoff-evaporation hydrological model for arid plain oasis has been used to investigate the evapotranspiration in Akesu plain (Hu et al., 2004; Tang et al., 2004). In order to simulate groundwater recharge and discharge, the model construction and code have been slightly modified as follows.
Farmland and canal system are the main recharge source. The canals are classified to four levels: trunk canal, branch canal, lateral canal and field canal. Water efficiency of canal system was identified by field experiments for each level canal in the typical irrigation area Awati (Fig.2). The results then are used in other irrigation area with regards to the actual situation. Groundwater recharge from canal loss (Rc) is estimated using the following equation:
where, is water efficiency of canal system, is canal seepage coefficient, describing the relationship between canal loss and resulting flux to groundwater in a lumped way, and V[m3] is the canal abstraction volume. Canal seepage coefficient highly depends on canal lining, canal cross section, underlying soil properties and the depth from canal bottom to water table. Moghazi et al. (1997) performed a field investigation in sandy soil to determine and evaluate the water losses for three different types of canals (earthen-uncompacted, compacted bed, and lining with bituminize jute). They determined that water lost by seepage was 98.8%, 98.9%, and 94.7% total water lost for earthen-uncompacted, compacted bed, and lining with bituminize jute mats canals, respectively. But they expected that the high canal seepage coefficient was because of the coarse-textured soil of the field site and concluded that evaporation losses were considerable and should be included in studies of water losses from canals under the high atmospheric demands of arid regions. In consideration of evaporation from shelter belt along canals, a canal seepage coefficient of 80% is used in this study.Groundwater recharge from farmland infiltration (Rf) is simulated by using the field soil module in the runoff-evaporation hydrological model mentioned above. Groundwater might supply field soil and crops when soil moisture is low, that usually occurs in spring and winter due to lack of irrigation at that time. All the water table depth in field is set equal in a lumped way.
A method modified from the DRAIN package of MODFLOW (McDonald et al., 1988) is used to estimate surface drainage from irrigation area. The groundwater discharge can be calculated as:
where, C[1/T] is drainage coefficient, describing water table decrease ratio due to head difference between water table and drainage level in a lumped way, Ad[M2] is the drainage area, h[M] is the phreatic level and hd[M] is the drainage level. The drainage coefficient highly depends on underlying soil conductivity and the interspace between offtakes. The drainage coefficient can take local adaptations into account and give a parameter to describe the drainage system development.Evaporation from phreatic water for bare soil land can be calculated as (Gardner, 1958; Mao et al., 1999; van Bavel et al., 1976):
where, E0[mm/T] is water surface evaporation, H is the water table depth, ,, are soil parameters. The method then can be extended to estimate evaporation from phreatic water with plants by using (H - r) instead of H, where r is effective rooting depth (Mao et al., 1999).The groundwater level is then calculated from the water store change in the aquifer. The calibration of the model is based on comparison of observed and calculated water table depths and drainage.