Detecting the sensitive area of evaporation duct and evaluating its impact on forecasts of electromagnetic propagation over the South China Sea

Evaporation ducts can significantly modify the atmospheric refractive index profile, leading to errors in electromagnetic (EM) propagation prediction, thereby affecting the reliability of maritime radio communication and radar detection. However, due to the uncertainty of initial conditions, it remains challenging to achieve high-precision forecasts of evaporation ducts. Detecting the sensitive areas of forecast errors and designing optimal observation strategies are proved to be effective techniques to reduce initial-condition uncertainty and improve the prediction skill of evaporation ducts. In this study, the nonlinear local Lyapunov exponent (NLLE) method is introduced to detect the sensitive areas of forecast errors for summer evaporation duct heights (EDHs) over the South China Sea (SCS). Results show that the sensitive areas are mainly distributed in the northern SCS, particularly along the coastal regions of China and Vietnam. The presence of these sensitive areas has a negative impact on the accuracy of EM propagation path loss (PL) forecasts. After excluding the sensitive areas, both the mean absolute error (MAE) and root mean square error (RMSE) of PL decreased, indicating that the detected regions exert a substantial influence on forecast skill. The improvement in PL forecasting was minor in 2024, whereas it was significant in 2025, suggesting that the effectiveness of the identified sensitive areas may vary under different environmental conditions. The results of these numerical experiments indicate that the detected sensitive areas are robust and reliable. This preliminary exploration provides encouraging results and highlights the potential for improving forecast performance. It is therefore worthwhile conducting further research on the sensitive areas of evaporation ducts in future work.