Hydroclimatic modelling of upper indus basin rivers predictability

• Published in: Modeling earth systems and environment 2024-02, Vol.10 (1), p.483-495
• Main Authors: Nawaz, Faisal, Hassan, Syed Ahmad
• Format: Article
• Language: English
• Subjects: Chemistry and Earth Sciences; Climate change; Climate prediction; Computer Science; Cost analysis; Disasters; Drought; Earth and Environmental Science; Earth Sciences; Earth System Sciences; Ecosystems; Environment; Evapotranspiration; Math. Appl. in Environmental Science; Mathematical Applications in the Physical Sciences; Natural disasters; Original Article; Physics; Precipitation; River flow; Rivers; Runoff; Spectrum analysis; Statistics for Engineering; Stream flow; Time series; Water resources
• ISSN: 2363-6203; 2363-6211
• DOI: 10.1007/s40808-023-01785-4

Climate change is one of the main factors affecting the habitats and water resources of the country. These changes may sometimes create natural disasters like floods and droughts around the world have done huge damages to the Pakistan in recent decades. Universal climatic variables such as temperature and precipitation influence runoff, river flow, evapotranspiration etc. Therefore, various techniques and models for the analysis and simulation of hydroclimatic time series have been projected. To minimize time and cost of the analysis, a nonparametric singular spectrum analysis (SSA) method was used to predict hydroclimatic variables. The SSA method has proven to be an influential tool for hydroclimatic data to examine important information about constrained components and further analysis. It is one of the smooth time series methods that does not require any prior assumption, such as the stationarity of the series or the normality of the residuals. This paper briefly explains the main steps of the technique and performs an SSA output to calibrate and validate the monthly temperature, precipitation and river flow for the Upper Indus Basin (UIB) rivers of ​​Pakistan. The SSA prediction and forecasting results are compared to known parametric techniques of multiple linear regression (MLR) and vector autoregression (VAR) methods. It can be stated that SSA leads to better results for both the calibration period and the validation period.