Understanding urban food deserts from a geo-data driven perspective

The concept of urban food deserts has been extensively analyzed and documented in developed countries, but in the Global South, it has not. Since developing countries have a distinct retail sector composition, its necessary to analyze and rethink the way to understand this concept. In developing countries and, specifically, in Peru, 70% of the retail channel are from small traditional retailers called nanostores, which complicate the last-mile logistics. In this sense, this paper analyzes the case of Lima, Peru regarding its fruit and vegetables logistics distribution system from a geodata-driven perspective by using 1) Geographical Information Systems (GIS) to theoretically identify urban food desert zones and 2) a clustering technique to obtain the different nanostores profiles and assess the validity of point 1. Furthermore, in this case of study, it was inferred that the healthy food supply is related to specific nanostores characteristics. Finally, the urban food desert conceptualization in current literature does not apply for urban areas in the case of Lima, Peru.