ABSTRACT
Effective management of road assets requires timely and accurate mapping and monitoring of road network conditions, which is currently limited by ground-level data collection methodologies that are labourious and costly, especially in developing countries. Remote sensing offers a promising alternative due to wide-area observation capabilities of artificial satellites and advancements in their spatial and temporal resolution technologies. This study explores the use of the free to access Copernicus Sentinel-2 satellite data focusing on the relationship between Sentinel-2 reflectance values and international roughness index (IRI) measured on the Kenya road pavement network. Fuzzy clustering identified at least three distinctions in spectral signatures that correspond to varying levels of continuous IRI. The Kruskal-Wallis and post-hoc Dunn’s tests confirmed statistically significant relationships between reflectance and IRI (p < 0.05 for each test), demonstrating that Sentinel-2 data can be a cost-effective tool for supporting proactive maintenance strategies particularly in resource constrained environments.
