Assessment of precipitation variability and uncertainty of stream flow in the Hindu Kush Himalayan and Karakoram River basins of Pakistan

The agricultural production system and hydropower production of Pakistan is largely dependent on melting water from the Hindu Kush Himalayan and the Karakoram River basins. Regardless of such significant prominence, a complete evaluation of the prevailing state of hydro-climatic variables is missing. In this context, we examine the precipitation variability and uncertainty of river flows based on diversity indexes and Shannon information entropy theory. The results indicate that the Shannon diversity index presents a clear depiction of precipitation variability on all metrological stations as compared to the Simpson diversity index. The maximum precipitation variability was found at Chilas station as compared to Hunza, Bunji and Gilgit stations. The results also show that the diversity indexes are highly, negatively correlated with standard deviation, and the amount of precipitation is less than 50 mm (dry months). At the decadal scale, the Hunza and Astore stations show higher precipitation evenness as compared to the Skardu, Bunji and Gilgit stations. The uncertainty analysis shows higher entropy value E_D = 90% for Indus River gauged at Mangla station. The higher entropy means the greater the chaos of the variables and the lower their certainty. The analysis exhibited that the rivers with high stream flow variability also show low entropy of its distribution and therefore higher stream flow concentration in the annual cycle.