Application of Semi-Distributed Model for Hydrological Assessment under Changing Climate – A Case Study of Panjkora River Catchment

Abstract

The accuracy of daily stream flow predictions in watershed is an immense challenge for un-gauged area particularly to changing climate due to quality of climate records. In current study, various gridded meteorological datasets of Asian Precipitation Highly Resolved Observational Data Integration Towards Evaluation (APHRODITE), Tropical Rainfall Measuring Mission (TRMM), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) and Climate Forecast System Reanalysis (CSFR) were used. Soil and Water Assessment Tool (SWAT) was applied to simulate and project the streamflows for Panjkora River watershed which is located in North-West of Pakistan. SWAT after calibration depicted better result with APHRODITE for precipitation and CFSR for minimum and maximum temperatures. The monthly simulated streamflows were calibrated from 2006-2009 (4 years) and validated for 2010-2012 (3 years). The model performance was evaluated using Coefficient of Determination (R2) and Nash-Sutcliffe Simulation Coefficient (NS), showed acceptable values of 0.70 and 0.58 for calibration period and 0.71 and 0.67 for validation period respectively, both laying within acceptable ranges. Further, validated model was used to study projected streamflows end of 21st century using bias corrected Himalayan Adaptation, Water and Resilience (HI-AWARE) General Circulation Models (GCMs) dataset of both Representative Concentration Pathways i.e. RCP4.5 and RCP8.5 scenarios, showing high and medium scenarios. Daily minimum temperatures (Tmin) are projected to experience a general increase for the coming years. 2025s experience highest increase in the summers with an increase viii of 1.9°C with RCP 4.5. Winter and spring have an increase of 1.3°C and 0.9°C respectively with RCP 8.5. For 2055s, summer again has the highest increase with 3.3°C, 2.7°C for winter, and 2.4°C for spring, all with RCP 8.5. For the last part of this century, with an increase of 5.4°C, summer again has the highest increase. Spring and winter have an increase of 4.4°C and 4.3°C respectively, all with RCP 8.5. Daily maximum temperatures (Tmax) are projected to follow a similar trend, with the maximum increase of 2.0°C in the autumn of 2025s for RCP 4.5, second to an increase of 1.2°C for spring and 0.5°C for the winter of the same time period with RCP 8.5. Moreover for 2055s, spring season is expected to experience the maximum increase of 3.2°C with RCP8.5. Autumn season also had a substantial increase of 3.0°C with RCP 4.5 pushing winter season to third place with an increase of 2.6°C with RCP 8.5. Towards the end of this century, 2085s had similar case, with an increase of 5.6°C, 4.8°C, and 4.6°C for spring, winter and autumn season respectively, all for RCP 8.5. Currently the area experiences its highest precipitation in winter season with a monthly average of 107.8 mm. Moreover winter and autumn precipitation is expected to increase and a decline is projected for the remaining part of the year. In the near future i.e. 2025s, highest decrease will be in summer (31.7 mm) with RCP 8.5 and an increase of 22.4 mm in the winters. however projected changes in 2055s show a different trend, although maximum decrease (38.7 mm) stays in summer with RCP4.5, the maximum increase (22.6 mm) shifts in autumn with RCP 8.5. And by the end of 21st century, a decrease of 41.1 mm is expected in summer with RCP4.5 and an autumn increase of 29.2 mm with RCP 8.5. Forecast of monthly flows shows temporal shift in peak flows as the area currently experiences its peak in April but from 2055s ix onwards, it is projected to be shift in March. The winters of 2025s will experience the highest increase of 28.17 Cumecs with RCP 8.5 while the same reflects a decrease in summers as 9.41 Cumecs. During the 2055s, winter flows are expected to rise by 45.72 Cumecs in comparison to a summer and spring decrease of 14.11 Cumecs and 33.01 Cumecs respectively (RCP 8.5). 2085s reflect a similar scenario of flows, increasing up to 63.62 Cumecs for winters and a decrease of 55.93 Cumecs in spring (RCP 8.5).