Machine Learning-Based Surface Refractivity Prediction in Coastal and Inland Regions of West Africa: A Comparative Study

DOI

10.2991/978-94-6463-644-4_21How to use a DOI?

Keywords

Surface radio refractivity; Machine learning; Temperature; Pressure; Relative humidity; Surface net solar radiation; Wind speed; Precipitation; and Potential evaporation

Abstract

This conference paper investigates the application of machine learning models for predicting surface radio refractivity across diverse climatic zones in West Africa, focusing on four key locations: Kano, Freetown, Accra, and Niamey. Accurate radio refractivity prediction is critical for optimizing telecommunications and radar systems, as it directly affects radio wave propagation. Using the ERA5 reanalysis dataset, surface refractivity was computed from key atmospheric parameters such as surface net solar radiation, wind speed, precipitation, and potential evaporation. Three machine learning models; LightGBM, Random Forest, and Gated Recurrent Unit (GRU), were trained and optimized using grid search and early stopping techniques. LightGBM achieved the lowest errors across all locations, with an MSE of 1635.56 and MAE of 32.47, outperforming Random Forest and GRU by approximately 4% and 67%, respectively, particularly in semi-arid and stable regions, while GRU underperformed in all cases. The study identifies the most influential atmospheric parameters and highlights the superior performance of LightGBM for refractivity prediction across the region’s coastal and inland zones.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - K. C. Onawumi
AU  - A. T. Adediji
AU  - S. T. Ogunjo
PY  - 2025
DA  - 2025/02/04
TI  - Machine Learning-Based Surface Refractivity Prediction in Coastal and Inland Regions of West Africa: A Comparative Study
BT  - Proceedings of the 8th URSI-NG Annual Conference (URSI-NG 2024)
PB  - Atlantis Press
SP  - 213
EP  - 223
SN  - 2352-541X
UR  - https://doi.org/10.2991/978-94-6463-644-4_21
DO  - 10.2991/978-94-6463-644-4_21
ID  - Onawumi2025
ER  -