Projection of hydrological drought based on SSP scenarios using surface water supply index and SWAT model in mountainous watershed

Abstract

The objective of this research is to explore climate change impact on hydrological drought in mountain watersheds. The Coupled Model Intercomparison Project Phase 6 (CMIP6) models and the Soil and Water Assessment Tool (SWAT) runoff simulation model were utilized to attain the objective of the study. Moreover, the surface water supply index (SWSI) was employed to examine hydrological drought. To correct the bias of the selected CMIP6 models (ACCESS-CM2, CNRM-CM6-1, CNRM-ESM2-1, INM-CM4-8, INM-CM5-0, MIROC6, and MIROC-ES2L), the historical period from 1990 to 2014 was used as a baseline. It was found that the MIROC-ES2L model exhibited greater accuracy compared to the other chosen models in the context of the examined mountain basin, namely the Ekbatan watershed in Iran. The calibration and validation of the SWAT model were conducted using the Sequential Uncertainty Fitting version 2 (SUFI-2) algorithm, covering the calibration period from 2004 to 2017 and the validation period from 2018 to 2020. The results indicated that the model effectively simulates runoff in the watershed. The highest percentage error, as measured by the Normalized Root Mean Square Error (NRMSE), was 18.78% during the validation phase. Results from a trend analysis of runoff using the Mann–Kendall test for the projected period (2023–2042) indicate that runoff is not expected to exhibit a significant decreasing or increasing trend during this timeframe. Furthermore, an assessment of runoff uncertainty within the projected period (2023–2042) revealed that the most substantial variability or uncertainty is associated with the month of April, with a range of 0 to 9.55 m³/s. The comparison of future runoff (2023–2042) with past runoff (2000–2020) indicates an increase in runoff during typically dry months (August, September, and October). This increase is particularly pronounced in September, where, under the SSP2-4.5 scenario and the best-performing model (MIROC-ES2L), runoff is projected to be nearly twelve times higher. This phenomenon may be attributed to the potential occurrence of intense rainfall events influenced by climate change during the dry season. Additionally, due to the generally low streamflow in this period, the relative increase appears more significant. When comparing drought conditions between the periods of 2023–2042 and 2000–2020, as assessed by SWSI, the results show that, on a 12-month scale, drought conditions in the typically wet months (February, March, April, and May) are expected to intensify under all scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5). In other words, these months are projected to become drier in the future. The most significant increase in drought severity is anticipated in February under the SSP5-8.5 scenario, reaching a magnitude of 86.84%. Overall, results indicated an intensification of hydrological drought during the wet months in the mountainous basin.