Comment on nhess-2022-244

Abstract

Hydrological drought forecasting can mitigate the socio-economic and ecological impacts of drought. It is an important disaster reduction strategy to forecast the occurrence of hydrological drought according to the forecasting system. In this paper, the conditional distribution model with human activity factor as exogenous variable was constructed to forecast the hydrological drought based on meteorological drought, and then compared with the traditional normal distribution model and conditional distribution model. The results show that the runoff series of Luanhe River Basin from 1961 to 2010 was non-stationary. For the traditional conditional probability models, the transition probabilities of drought were affected by SPI time scales and forecasting periods. In order to analyze the impact of human activities on hydrological drought, we constructed the human activity factor based on the method of restoration. Subsequently, the conditional distribution models involving human index were constructed and the influence of human activities on drought transition probability was analyzed. With the increase of human index (HI) value, hydrological droughts tend to transition to more severe droughts. Finally, a scoring mechanism was applied to evaluate the performance of three drought forecasting models. According to the scores of the three drought forecasting models, the conditional distribution model involving of human activity factor can further improve the forecasting accuracy of drought in Luanhe River Basin.