Spatiotemporal precipitation trends and associated large-scale teleconnections in northern Pakistan

The effects of climate change are unparalleled in magnitude,ranging from changing weather patterns that endanger food production to increasing sea levels that increase the likelihood of catastrophic flooding.Therefore,the extent of such variations on regional and local is imperative.We used monthly precipitation data from 25 meteorological stations in northern Pakistan(NP)to document the observed changes in seasonal and annual precipitation.The station density in the NP is small and unevenly distributed;therefore,ERA-5 reanalysis data are used to supplement the observed dataset to assess the spatial trends in NP.The non-parametric Mann-Kendall(MK),Sen's Slope estimator(SSE),and Sequential Mann-Kendall(SQMK)tests were performed to assess the trends.In addition,the wavelet analysis technique was used to determine the association of precipitation with various oceanic indices from 1960 to 2016.Results indicate that maximum precipitation was shown in the annual and summer seasons.In NP,annual,winter,spring,and summer precipitation declined,while an increase in autumn was observed at a rate of 0.43 mm/decade between 1989 and 2016.The spatial trends for observed and ERA-5 reanalysis datasets are almost similar in winter,spring,and autumn;however,some disagreement is observed in both datasets in the summer and annual precipitation trends in NP during 1960-2016.Between 1989 and 2016,summer and annual precipitation increased significantly in Region III.However,seasonal and annual precipitation decreased in NP between 1960 and 2016.Moreover no prominent trends in annual precipitation until the mid-1980s,but an apparent increase from 1985 onwards.Annual precipitation has increased in all elevations except at the 500 m-1000 m zone.ENSO(El Niño-Southern Oscillation)shared notable interannual coherences among all indices above 16-64 months.Inter-decadal coherence with ENSO,AO(Arctic Oscillation),and PDO(Pacific Decadal Oscillation)in NP for 128 months and above.Generally,AO,AMO(Atlantic Multidecadal Oscillation),and NAO(North Atlantic Oscillation)exhibit less coherence with precipitation in NP.The regression of seasonal and annual precipitation reveals that winter and spring precipitation have higher linear regression with AO and ENSO,respectively,while both AO and ENSO are also dominated at the annual scale.Similarly,the IOD,and PDO indices have a higher influence in summer precipitation.The findings could help water resource managers and climate researchers develop a contingency plan for better water resource management policies in the face of changing climate change in Pakistan,particularly in the NP.