Simulating the impact of climate and land cover changes on river discharge in the Maros watershed

Abstract

The Maros watershed is a rapidly urbanizing tropical karst region experiencing severe land degradation and annual flooding. To address these extreme hydrological risks, this study proposes an integrated climate-land cover dynamic simulation framework utilizing the soil and water assessment tool (SWAT) to quantify the impact of each driver on extreme discharge dynamics. According to the results of the analysis of precipitation and temperatures from 1983–2023, there was a fluctuating trend in annual precipitation, with an average of 1,979.53 mm/year (the highest in January, with a value of 370.16 mm/month), while the temperature was 26.27 °C. However, the results of the 2064 projection show that the trend of rainfall increases in rainy months by between 1% and 4% and decreases in dry months by 32%. In the land cover area projected by 2064, land degradation occurs, where the area of secondary forests is projected to decrease by 50.41 ha/year (382.18 ha/year from 2014–2023), as well as the expansion of agricultural land (26.25 ha/year for agriculture/crop land and 0.54 ha/year for paddy areas) and settlements (24.11 ha/year). The results of the SWAT simulation describe a highly sensitive (striking) flow characteristic where extreme intense rains occurred on December 7, 2021, resulting in simulated maximum flows of 1,206 m3/s in sub watershed 6, 618.70 m3/s in sub watershed 23 and 289.80 m3/s in sub watershed 46. In terms of the projection, the maximum flow occurred on February 13, 2064, with flows of 184.10 m3/s in sub watershed 6, 133 m3/s in sub watershed 23, and 79.40 m3/s in sub watershed 46. The resulting flow value represents a rapid response between the extreme precipitation event and land cover change for the discharge surge. The results of the validation of the best SWAT model at the Lekopaccing Dam revealed satisfactory performance, with a Nash-Sutcliffe (NSE) value of 0.75.