作者简介:吴怿昊，男，1987年生，副教授，硕士生导师，主要从事重力数据融合、建模和卫星测高数据处理方面的研究. E-mail: yihaowu@hhu.edu.cn

通讯作者: 罗志才，男，1966年生，教授，博士生导师，现主要从事物理大地测量学和卫星重力学等方面的研究. E-mail: zcluo@hust.edu.cn

中图分类号: P223

This research studies the method for high-resolution regional gravity field modeling from airborne gravity gradient tensor (GGT) based on radial basis functions. Considering the heavy computational load and numerical inefficiency of the traditionally used Legendre polynomial expansion, a fast modeling approach for regional gravity field recovery from the full GGT is proposed. We use Poisson wavelets as an example, and the analytical expressions of Poisson wavelets and the associated first- and second-order partial derivatives are recursively deduced, to link the full GGT to the representation of regional gravity field. By combining the full airborne GGT, we model the regional gravity field model with a resolution of 0.5 km over the St. George′s Bay in Canada, and quantify the effects on the gravity field recovery using individual components of GGT. The results show that the solution computed from the vertical component of airborne GGT provides the best quality when a single component is used. Moreover, the incorporation of multiple components of airborne GGT magnifies the gravity anomalies and further improves the quality of the local gravity field, compared with the solutions computed using single components. By combining the full airborne GGT, the derived regional gravity field model has improved accuracy in short wavelengths, compared to the recently released high-degree global geopotential models and satellite altimetric gravity data.