Periurban areas in mega-cities tend to constantly change their land use due to their urbanization growth. This chapter evaluates periurban dynamics and agricultural traditionality in a case study south of Mexico City. Sentinel-2 satellite imagery and the Random Forest classifier (RF) were used in combination with participatory mapping and ethnographic fieldwork to map three indigenous towns with different predominance of crops and urban spatial patterns. Within the study area, six cover type classes were delimited: Urban, Forest, Scrubland, Induced Grassland, and Maize, including Milpa, and Nopal for two time periods, 2016 and 2022. The classification was performed using RF. Overall accuracy was high for the training group of 99.7% (k = 0.996) for 2016 and 99.6% (k = 0.99) for 2022. The results for the validation group were also high, with an overall accuracy of 84.3% (k = 0.81) for 2016 and 81.0% (k = 0.77) for 2022. The highest accuracy in the cover classifications was obtained using b2, b3, b4, and b8 spectral bands with 10 m spatial resolution. In two cases, the Crop cover had the greatest change while it was the Urban cover in the most urbanized town. The Nopal cover had the greatest negative loss, unlike the town with the most agricultural traditionality, which was the Grass cover. Through participatory mapping, interviews, and fieldwork, it was found that producers with greater crop management knowledge and less urbanization maintain much more agricultural traditionality. This research reveals that agricultural traditionality is a positive factor that restricts land use change between towns in this periurban area of Mexico City. The discussion argues that the development planning strategies that lead the transition of periurban areas toward sustainability should consider the spatial patterns of agricultural traditionality change in connection with urban landscapes, as part of an adaptive governance.
Elsevier, Modern Cartography Series, Volume 11, 2024, Pages 663-685
