Publications by authors named "Ibrahim Ademola Adeniran"
Article Synopsis
- Near-surface air temperature (Tair) is essential for understanding urban heat and its effects on health, but traditional estimation methods often overlook the spatial differences in temperature.
- This study introduces a federated learning artificial neural network (FLANN) framework that uses comprehensive thermal data from multiple satellite sources and weather stations to improve Tair prediction.
- Compared to existing models, FLANN demonstrated significantly better accuracy with a high Pearson correlation coefficient (r = 0.98) and a low root mean square error (RMSE = 0.97 K), making it particularly effective for analyzing urban heat islands in cities like Hong Kong.
