Method For Extracting Stratospheric Gravity Waves Based On Deep Learning

As an important form of atmospheric motion,wave is one of the key processes in atmospheric dynamics.The motion of each scale in the atmosphere and its associated weather system process are all generated by certain atmospheric fluctuations.At present,atmospheric science has deeply studied the mechanism and dynamic characteristics of atmospheric fluctuations.Gravity wave(GW),which is one of the most significant middle and upper atmospheric dynamic processes,plays a key role in determining the global atmospheric circulation and is an important driving force for the middle atmospheric circulation and the transport of material and energy.Using various means to study the stratospheric gravity wave is helpful to deepen the understanding of the atmospheric environment of near space,which is an important strategic region,and is of great significance to the study of stratospheric atmospheric environment.Gravity wave potential energy(Ep)is an important parameter that characterizes GW intensity,so understanding its global distribution is necessary.In this paper,the calculation methods of stratospheric gravitational wave potential energy and dominant vertical wavelength are introduced,and the temporal and spatial variation rules are studied.Then a deep learning algorithm(DeepLab V3+)is used to estimate the stratospheric GW Ep.The deep learning model inputs are ERA5 reanalysis datasets and GMTED2010 terrain data.The output is the estimated average GW Ep over 20～30 km from 60°S～60°N.The average GW Ep over 20～30 km calculated by COSMIC radio occultation(RO)data is used as the label of the model training.The results showed that(1)this method can effectively estimate the zonal trend of GW Ep.However,the errors between the estimated and measured value of Ep are larger in low-latitude regions than in mid-latitude regions because the measured Ep has errors associated with interpolation to the grid.Additionally,the error tends to be amplified in low-latitude regions because the GW Ep is larger and the RO data are relatively sparse,which affects the training accuracy.(2)The estimated Ep shows seasonal variations,which are stronger in the winter hemisphere and weaker in the summer hemisphere.(3)The effect of quasi-biennial oscillation(QBO)can be clearly observed in the monthly variation in the estimated GW Ep.At present,only the occultation data can be used to calculate the potential energy of gravity wave,but the COSMIC RO data,which is the highest horizontal resolution occultation data and covers the globe at present,only has a time span of more than ten years.The significance of this study is the verification of the feasibility of the deep learning method to estimate GW Ep.Abundant reanalysis data can be used to estimate stratospheric GW Ep in recent decades so that the long-term variation in stratospheric GWs in mid-and low-latitude regions can be determined.Additionally,for future studies,topographic data and numerical prediction can also be considered as inputs of the deep learning model to realize the prediction of GW Ep.