In the ANUSPLIN thinplate smooth spline interpolation, the accuracy of interpolation results is mainly determined by choosing the independent covariates. This article selected the weather processes with poor visibility in heavy fogs and hazes from 2017 to 2019, using 183 visibility observation sites to interpolate visibility, and introduced the Himawari8 satellite channel data and DEM data as covariables to improve the visibility interpolation results. The visibility interpolation effects are compared and analyzed. The results show that the visibility interpolation effect of Himawari8 data and DEM data as covariables is significantly improved in accuracy, especially in the inversion of the boundary range and texture of fog and haze areas. The accuracy of interpolation using the covariate method and interpolation only using the observed values is greatly improved.