基于GAMLSS模型的渭河流域非平稳气象干旱评估

Assessment of Nonstationary Meteorological Drought in the Weihe River Basin Based on GAMLSS Model

[目的]识别气候变化影响下降水序列的非平稳特性,评估渭河流域气象干旱特征,为干旱管理及风险评估提供有效参考。[方法]基于GAMLSS模型构建以大尺度气候因子为协变量的非平稳标准化降水指数(NSPI),结合历史干旱事件和Copula概率分析,并与传统的标准化降水指数(SPI)进行了比较,讨论了其在渭河流域的适用性。[结果]大尺度气候因子对降水序列的影响具有季节性,AOI,SOI和ENSO指数对研究区降水影响最广泛,并且以气候因子为协变量的非平稳模型拟合效果在各月份均优于平稳模型。结合游程理论和历史干旱事件,发现NSPI在干旱事件识别中表现更好,NSPI识别的特旱事件的发生频率比SPI高。此外,SPI和NSPI的干旱特征变量的二维联合分布的最优Copula函数分别为Frank和Gumbel,通过对比联合概率和重现期,发现SPI低估了干旱发生的风险。[结论]气候变化背景下,考虑非平稳特性的NSPI表现出了更准确的干旱特征评估效果。

[Objective]The aims of this study are to identify the non-stationary characteristics of the descending water series affected by climate change,to evaluate the meteorological drought characteristics in the Weihe River Basin,so as to provide a more effective reference for drought management and risk assessment.[Methods]Based on GAMLSS model,a non-stationary standardized precipitation index(NSPI)with large-scale climate factors as covariables was constructed.The applicability of NSPI in the Weihe River Basin was discussed by combining with historical drought events and Copula probability analysis,and comparing with the traditional standardized precipitation index(SPI).[Results]The influence of large-scale climate factors on precipitation series was seasonal.AOI,SOI and ENSO indexes had the most extensive influence on precipitation in the study area.Moreover,the fitting effect of non-stationary model with climate factors as covariable is better than that of stationary model in each month.Combining the run course theory and historical drought events,it is found that NSPI performs better in drought event recognition,and the occurrence frequency of extreme drought events identified by NSPI is higher than that by SPI.In addition,the optimal Copula function of the two-dimensional joint distribution of drought characteristic variables of SPI and NSPI is Frank and Gumbel,respectively.By comparing the joint probability and recurrence period,it is found that SPI underestimates the risk of drought occurrence.[Conclusion]In the context of climate change,NSPI considering non-stationary characteristics shows a more accurate evaluation of drought characteristics.