Estimation of two reanalysis datasets on the application to atmospheric refractive environment over the global Sea

Atmospheric refractivity over the sea is one of important factors which affects the electromagnetic wave propagation, and is also a basic element which comprises dataset of electromagnetic environment. Since reanalysis data is an indispensable data source to study the atmospheric refractivity, comparative analysis of two kinds of reanalysis datasets, including ERA-I data and reanalysis data (FNL) provided by European Centre for Medium-Range Weather Forecasts (ECMWF) and US National Centers for Environmental Prediction (NCEP) respectively, are performed to analyze the atmospheric refractive environment based on the GPS sounding data in the direct application and numerical simulation as the initial field. The results show that the FNL dataset is better in the open sea facing the ocean, and the root mean square errors of the modified atmospheric refractivity from the FNL data are 4.7 and 2.8 M units in the Arctic and Antarctic respectively．However, in continental or marginal sea, ERA-I reanalysis data depict atmospheric refractive index more accurate and ERA-I data can identify most ducting cases against FNL data especially in the sea. For instance in the Middle East where ducting events are highly prevalent, the correct rates of ducting identified by ERA-I and FNL reanalysis data are 58.1% and 45.4%. On the numerical simulations, the results produced by WRF model with two datasets as initial fields have almost the same errors, and the overall root mean square errors of the modified atmospheric refractivity at each station are less than 1 M units. This results provide a reference for the application of atmospheric refractive environment in different sea areas.