Improving AVSWAT Stream Flow Simulation by Incorporating Groundwater Recharge Prediction in the Upstream Lesti Watershed, East Java, Indonesia

摘要

The upstream Lesti watershed is one of the major watersheds of East Java in Indonesia, covering about 38093 hectares. Although there are enough water resources to meet current demands in the basin, many challenges including high spatial and temporal variability in precipitation from year to year exist. It is essential to understand how the climatic condition affects Lesti River stream flow in each sub basin. This study investigated the applicability of using the Soil and Water Assessment Tool (SWAT) with the incorporation of groundwater recharge prediction in stream flow simulation in the upstream Lesti watershed. Four observation wells in the upstream Lesti watershed were used to evaluate the seasonal and annual variations in the water level and estimate the groundwater recharge in the deep aquifer. The results show that annual water level rise was within the 2800 - 5700 mm range in 2007, 3900 - 4700 mm in 2008, 3200 - 5100 mm in 2009, and 2800 - 4600 mm in 2010. Based on the specific yield and the measured water level rise, the area-weighted groundwater predictions at the watershed outlet are 736, 820.9, 786.7, 306 4 mm in 2007, 2008, 2009, and 2010, respectively. The consistency test reveals that the R-square statistical value is greater than 0.7, and the DV (%) ranged from 32 - 55.3% in 2007 - 2010. Overall, the SWAT model performs better in the wet season flow simulation than the dry season. It is suggested that the SWAT model needs to be improved for stream flow simulation in tropical regions.