Detection of open-air weekly markets using remote sensing techniques

Open-air markets play a vital role in food security, nutrition, and local economies, yet their spatiotemporal dynamics remain poorly understood. This study explores the use of Sentinel-1 SAR data to detect open-air markets operating weekly by identifying consistent changes in backscatter properties. The method was tested in regions with known market locations, including San Francisco el Alto, Guatemala, and Montpellier, France, where it successfully identified market activity and related patterns. It was then scaled to a national level in Rwanda by limiting the analysis to bare soil and built-up areas using Google Dynamic World V1 land cover data. The results were promising, identifying several markets and providing a foundation for further refinement. However, the approach requires improvements to differentiate markets from other areas with weekly patterns, such as parking lots and construction sites, and thorough validation across diverse contexts. This study demonstrates the potential of remote sensing to characterize open-air markets and their temporal dynamics, contributing valuable data to food system models in data-scarce regions.