This study aimed to examine the spatiotemporal seasonal and annual trends of rainfall indices in Perak, Malaysia, during the last 35 years, as any seasonal or spatial variability in rainfall may influence the regional hydrological cycle and water resources. Mann–Kendall and Sequential Mann–Kendall (SMK) tests were used to assess seasonal and annual trends. Precipitation concentration index was used to estimate variations in rainfall concentration, and Theil–Sen’s slope estimator was used to determine the spatial variability of rainfall. It was found that most of the rainfall indices are showing decreasing trends, and it was most prominent for the southwest monsoon season with a decreasing rate of 2.20 mm/year. The long-term trends for seasonal rainfall showed that rainfall declined by 0.29 mm/year during the southwest monsoon. In contrast, the northeast and the inter-monsoon seasons showed slight increases. Rainfall decreased gradually from 1994 to 2008, and the trend became more pronounced in 2008. On a spatial basis, rainfall trends have shifted from the western regions (i.e., −19 mm/year) to the southeastern regions (i.e., 10 mm/year). Overall, slightly decreasing trends in rainfall were observed in Perak Malaysia.

A major part of Pakistan’s economy is dependent upon agriculture which is irrigated from the water resources of the Upper Indus Basin (UIB). Therefore the human impact of hydroclimatic variability in this area is of paramount importance. The Upper Indus Basin is characterized by uncertain hydro-climatic behaviour with changing patterns in different sub-basins. Many studies have worked on hydro-climatic trends at basin scale but only few studies focused on the hydroclimate, precipitation dynamics and their magnitude at sub-basin level. Based upon this scenario, high resolution seasonal and annual climatology of UIB was developed. It is based on precipitation normals 1995-2017 obtained from four different gridded satellite datasets (Aphrodite, Chirps, PERSIANN-CDR and GPCC) as well as quality- controlled high and mid elevation ground observations (1250–4500 m a.s.l.). The quality-control of the gridded dataset is computed by the anomaly method. In order to, evaluate the data quality of the gridded rainfall, four statistics i.e., BIAS, Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) are used in this study. Using running trends and spectral analysis with multi-gauge based anomaly, the study analyses the precipitation and runoff   seasonal and annual temporal variability at sub-basin scale. For this, Mann–Kendall test was employed to detect the presence of any trend while their slope is calculated by Theil Sen’s slope method. The nonparametric Pettitt Test was also used in this study to eventually identify the change point in hydro-climatic time series. The results indicated that bias corrected CHIRPS precipitation datasets performed better in simulating precipitation with RMSE, MAE, MAPE [%] and BIAS followed by APHRODITE. The annual and seasonal precipitation climatology exhibited higher precipitation in the lower side of the basin. The comparison between short and long duration climatologies is being investigated as well. The annual running trend analysis of precipitation exhibited a very slight change whereas a more significant increase was found in the winter season (DJF) and most of sub-basins feature a significant decreasing rate in precipitation and constant change point within the monsoon period (JJA). Similarly, trend analysis for runoff in main rivers of Upper Indus Basin at Gilgat, Indus (Besham Qila, Bunji) exhibit nonsignificant increase except Hunza and Indus at Kharmong which are showed decrease annual trends and will be further investigated for seasonal patterns. Overall, these findings would assist to better understand precipitation, snow- and ice-melt runoff dynamics, addressing the hydroclimatic behaviour of the Karakoram region.

A high resolution seasonal and annual precipitation climatology of the Upper Indus Basin was developed, based on 1995-2017 precipitation normals obtained from four different gridded datasets (Aphrodite, CHIRPS, PERSIANN-CDR and ERA5) and quality-controlled high and mid elevation ground observations. Monthly precipitation values were estimated through the anomaly method at the catchment scale and compared with runoff data (1975-2017) for verification and detection of changes in the hydrological cycle. The gridded dataset is then analysed using running trends and spectral analysis and the Mann–Kendall test was employed to detect significant trends. The nonparametric Pettitt test was also used to identify the change point in precipitation and runoff time series. The results indicated that bias corrected CHIRPS precipitation dataset, followed by ERA5, performed better in terms of RMSE, MAE, MAPE and BIAS in simulating rain gauge-observed precipitation. The running trend analysis of annual precipitation exhibited a very slight increase whereas a more significant increase was found in the winter season (DJF). A runoff coefficient value greater than one, especially in glacierized catchments (Shigar, Shyok and Gilgit) indicate that precipitation was likely underestimated and glacial melt in a warming climate provides excess runoff volumes. As far as the streamflow is concerned, variabilities are more pronounced at the seasonal rather than at the annual scale. At the annual scale, trend analysis of discharge shows slightly significant increasing trend for the Indus River at the downstream Kachura, Shyok and Gilgit stations. Seasonal flow analysis reveals more complex regimes and its comparison with the variability of precipitation favours a deeper understanding of precipitation, snow- and ice-melt runoff dynamics, addressing the hydroclimatic behaviour of the Karakoram region.