Multiple–linear-regression (MLR) impact radius diagnosis

The impact radius of wind observations is diagnosed using a multiple–linear-regression-based framework that quantifies the spatial decay characteristics of observational influence on the NWP model initial fields. Specifically, wind lidar network observations are combined with the model background fields through multivariate linear regression, and the distance-dependent decay behavior of the regression weighting coefficients is used to determine the effective impact radius of the meridional and zonal wind components.