Comparison between canonical vine copulas and a meta-Gaussian model for forecasting agricultural drought over China

[摘要] Agricultural drought mainly stems from reduced soil moisture andprecipitation, and it causes adverse impacts on the growth of crops andvegetation, thereby affecting agricultural production and food security. In order to develop drought mitigation measures, reliable agricultural droughtforecasting is essential. In this study, we developed an agriculturaldrought forecasting model based on canonical vine copulas inthree dimensions (3C-vine model) in which antecedent meteorologicaldrought and agricultural drought persistence were utilized as predictors.Furthermore, a meta-Gaussian (MG) model was selected as a reference toevaluate the forecast skill. The agricultural drought in China in August of 2018 wasselected as a typical case study, and the spatial patterns of 1- to 3-monthlead forecasts of agricultural drought utilizing the 3C-vine model resembledthe corresponding observations, indicating the good predictive ability ofthe model. The performance metrics – the Nash–Sutcliffe efficiency (NSE), the coefficient ofdetermination ( R 2 ), and the root-mean-square error (RMSE) – showed that the3C-vine model outperformed the MG model with respect to forecasting agricultural droughtin August for diverse lead times. Moreover, the 3C-vine model exhibitedexcellent forecast skill with respect to capturing the extreme agricultural drought overdifferent selected typical regions. This study may help to guide droughtearly warning, drought mitigation, and water resource scheduling.

[学科分类] 妇产科学