Distinct Interannual Variability and Physical Mechanisms of Snowfall Frequency over the Eurasian Continent during Autumn and Winter

This study investigates the dominant modes of interannual variability of snowfall frequency over the Eurasian continent during autumn and winter,and explores the underlying physical mechanisms.The first EOF mode(EOF1)of snowfall frequency during autumn is mainly characterized by positive anomalies over the Central Siberian Plateau(CSP)and Europe,with opposite anomalies over Central Asia(CA).EOF1 during winter is characterized by positive anomalies in Siberia and negative anomalies in Europe and East Asia(EA).During autumn,EOF1 is associated with the anomalous sea ice in the Kara–Laptev seas(KLS)and sea surface temperature(SST)over the North Atlantic.Increased sea ice in the KLS may cause an increase in the meridional air temperature gradient,resulting in increased synoptic-scale wave activity,thereby inducing increased snowfall frequency over Europe and the CSP.Anomalous increases of both sea ice in the KLS and SST in the North Atlantic may stimulate downstream propagation of Rossby waves and induce an anomalous high in CA corresponding to decreased snowfall frequency.In contrast,EOF1 is mainly affected by the anomalous atmospheric circulation during winter.In the positive phase of the North Atlantic Oscillation(NAO),an anomalous deep cold low(warm high)occurs over Siberia(Europe)leading to increased(decreased)snowfall frequency over Siberia(Europe).The synoptic-scale wave activity excited by the positive NAO can induce downstream Rossby wave propagation and contribute to an anomalous high and descending motion over EA,which may inhibit snowfall.The NAO in winter may be modulated by the Indian Ocean dipole and sea ice in the Barents-Kara-Laptev Seas in autumn.

Application of the improved dung beetle optimizer,muti-head attention and hybrid deep learning algorithms to groundwater depth prediction in the Ningxia area,China

Due to the lack of accurate data and complex parameterization,the prediction of groundwater depth is a chal-lenge for numerical models.Machine learning can effectively solve this issue and has been proven useful in the prediction of groundwater depth in many areas.In this study,two new models are applied to the prediction of groundwater depth in the Ningxia area,China.The two models combine the improved dung beetle optimizer(DBO)algorithm with two deep learning models:The Multi-head Attention-Convolution Neural Network-Long Short Term Memory networks(MH-CNN-LSTM)and the Multi-head Attention-Convolution Neural Network-Gated Recurrent Unit(MH-CNN-GRU).The models with DBO show better prediction performance,with larger R(correlation coefficient),RPD(residual prediction deviation),and lower RMSE(root-mean-square error).Com-pared with the models with the original DBO,the R and RPD of models with the improved DBO increase by over 1.5%,and the RMSE decreases by over 1.8%,indicating better prediction results.In addition,compared with the multiple linear regression model,a traditional statistical model,deep learning models have better prediction performance.