Simulating the effect of change in land cover and rainfall in Upper Citarum Watershed: calibration and sensitivity analysis of GenRiver model

Land use and climate change significantly affect watershed hydrological conditions and, hence, influence the effectiveness of a watershed in regulating landscape water balance. Ensuring that land allocation and utilization are managed sustainably is crucial and watershed management must be supported with integrated watershed planning. This study aimed to examine the ability of the Generic Riverflow (GenRiver) model to support watershed planning as a tool to assess the conditions of a watershed’s hydrologcal functions, particularly, to examine to what extent the watershed can buffer the impact of changes in land use and climate. We tested the model on an important watershed in West Java Province — the Upper Citarum — which is a part of one of the Citarum Watershed, a national priority. A process of model calibration and sensitivity analysis was conducted to determine the feasibility of GenRiver in simulating Upper Citarum conditions, in particular, to estimate the water balance at landscape level. The results of the GenRiver simulation using data from 2012–2016 showed that on average 37% of the rainfall in the watershed becomes surface flow (run-off), 7% becomes sub-surface flow and 20% baseflow. Sensitivity analysis was carried out by compiling five land-cover scenarios and three rainfall scenarios that were considered to represent various conditions, including extreme conditions such as 1) the entire area became open (extremely negative); and 2) the entire area became forest (extremely positive). The results of the extremely negative scenario show that a fully degraded condition of the watershed with a dominance of open land has the potential to increase surface runoff up to 70% of rainfall. Meanwhile, improvement of land cover in the watershed by reforestation (extremely positive scenario) would be able to reduce surface runoff by up to 20% of total rainfall. The model calibration and validation evaluation showed that GenRiver performed satisfactorily and was sensitive enough to capture the range of scenarios. Therefore, GenRiver can be used for creating policy-based scenarios that simulate the effect of land-use interventions or restoration programmes in Upper Citarum Watershed and beyond. A soil-erosion module should be added so that an even stronger policy recommendation can be developed