Uncertainty analysis for evaluating flow regime alteration of Jinsha River based on Indicators of Hydrologic Alteration

The uncertain relationship between data length and evaluation of hydrological alteration has perplexed researchers. In this study, the length of record (LOR) method is applied to analyse the uncertainty in the impact of different data lengths on the Indicators of Hydrologic Alteration (IHA) evaluation in the Jinsha River. The results indicate: (1) For the pre-impact period, a daily streamflow data length of 20 years should be required to meet all the indicators, but 5-10 years of data could be sufficient when analysing some specific indicators. (2) A logarithmic correlation was found between the LOR results and the coefficient of variation (CV) of the IHA statistics, which can be considered reference factors to reduce the computational complexity when we expect to obtain LOR results from other nearby hydrological stations. (3) For the post-impact period, using a data length of 10 years is sufficient to evaluate the impact of large reservoirs on the river hydrological regime, and the actual alteration of hydrological regime would be overestimated by using a shorter data length.