Utilization of WRF 3D Meteorological Data to Calculate Slant Total Delay for InSAR Atmospheric Correction

Abstract

The atmosphere introduces excess delays into the synthetic aperture radar (SAR) signal trajectory, especially in the troposphere. InSAR atmospheric correction methods include the use of SAR data and external water vapor products. The latter is more effective. However, since the removal of atmospheric effects should use atmospheric delay products in the direction of the line of sight (LOS), it is necessary to convert the zenith total delay to slant delay in the LOS direction. Conventionally, the zenith delay is divided by the cosine of the average incident angles to obtain slant phase delays. But this method could cause large errors because it ignores the atmospheric horizontal gradient change and the small-scale vertical structure. These problems can be solved by using three-dimensional atmospheric data simulated by numerical models, especially in the case of intense weather changes or complex terrain. However, few scholars paid attention to the application into InSAR atmospheric correction, because of the computation complexity and low efficiency. As the requirement for higher accuracy and the introduction of large errors caused by increasing incidence angles, it is significantly imperative to make the utmost of this method. Weather Research Forecast (WRF) model can provide the precipitate water vapor (PWV) and refraction index at different levels in the three dimensions, and then the slant total delay can be obtained for removing the atmospheric effect on the InSAR process. The results demonstrate that using 3D data can obtain more accurate slant total delay and improve the accuracy of surface deformation from InSAR technology.