文献类型：会议论文

英文题名：A method to calculate vertical temperature gradient and temperature advection based on flat-floating sounding data

作者：Chang, Shujie[1,2];Huang, Sixun[3];Li, Yongchi[1,2]

机构：[1]Guangdong Ocean Univ, Lab Coastal Ocean Variat & Disaster Predict, Dept Educ Guangdong Prov,Key Lab Climate Resource, Coll Ocean & Meteorol,South China Sea Inst Marine, Zhanjiang 524088, Peoples R China;[2]Minist Nat Resources, Key Lab Space Ocean Remote Sensing & Applicat, Beijing 100081, Peoples R China;[3]Natl Univ Def Technol, Coll Meteorol & Oceanol, Changsha 410073, Peoples R China

会议论文集：2nd International Conference on Applied Mathematics, Modelling, and Intelligent Computing (CAMMIC)

会议日期：MAR 25-27, 2022

会议地点：ELECTR NETWORK

语种：英文

外文关键词：numerical differentiation; Tikhonov regularization; temperature gradient; temperature advection

外文摘要：The vertical temperature gradient is an important indicator of atmospheric stratification, and the horizontal temperature gradient describes the variation in atmospheric temperature in a given horizontal direction. In the meanwhile, Temperature advection is a phenomenon whereby the temperature changes as a result of the horizontal air movements that plays an important role in the development of weather systems and weather phenomena. In this paper, a one-dimensional numerical differentiation algorithm is applied to calculate the temperature gradient and temperature advection from the temperature and wind fields obtained from flat-floating sounding data, and the results are compared with those of the central difference method. The comparison shows that the one-dimensional numerical differentiation algorithm is stable and feasible. However, in the calculation of the vertical temperature gradient, the advantages of the one-dimensional numerical differentiation algorithm are not apparent because of the high accuracy of the flat-floating sounding data. The relative error of the temperature-advection associated with the one-dimensional numerical differentiation algorithm is two orders of magnitude less than that associated with the central difference method. In addition, the relative error of the one-dimensional numerical differentiation algorithm is more stable. These results show that the issue of calculating partial derivatives based on observation data is ill-posed in mathematics and that the one-dimensional numerical differentiation algorithm is better suited to solve such issues than is the central difference method.