Fusion of Propagation Observations and Numerical Weather Prediction via Warping Functions

The inverse problem in trans-horizon electromagnetic (EM) propagation has been solved by mapping modeled refractive environments ( $\mathcal{M}$ ‘s) into the space of observed propagation d and selecting the best $\mathcal{M}$ based on goodness-of-fit. Up to the present, parametric ad hoc refractivity models have been used, e.g., [2]–[4]. We modify this approach by using an a priori $\mathcal{M}$ from numerical weather prediction (NWP) and solving for warping coefficients (v's) within a 2-dimensional variational (2DVAR) analysis.