Analyzing urban expansion and land use dynamics in Bagua Grande and Chachapoyas using cloud computing and predictive modeling

Urban growth and Land Use/Land Cover (LULC) changes have increased in recent decades due to anthropogenic activities. This study explored past and projected future LULC changes and urban growth patterns in the Bagua Grande and Chachapoyas districts using Landsat imagery, cloud computing, and predictive models for 1990 to 2031. The analysis of satellite images was grouped into four time periods (1990–2000, 2000–2011, 2011–2021 and 2021–2031). The Google Earth Engine (GEE) cloud-based system facilitated the classification of Landsat 5 ETM (1990, 2000, and 2011) and Landsat 8 OLI (2021) images using the Random Forest (RF) model. A simulation model integrating Cellular Automata (CA) and an Artificial Neural Network (ANN) Multilayer Perceptron (MLP) in the MOLUSCE plugin of QGIS was used to forecast urban sprawl to 2031. The resulting maps showed an overall accuracy (OA) of over 92%. A decrease in forested area was observed, from 20,807.97 ha in 1990 to 14,629.44 ha in 2021 in Bagua Grande and from 7,796.08 ha to 3,598.19 ha in Chachapoyas. In contrast, urban areas experienced a significant increase, from 287.49 to 1,128.77 ha in Bagua Grande and from 185.65 to 924.50 ha in Chachapoyas between 1990 and 2021. By 2031, the urban area of Bagua Grande is expected to increase from 1,128.77 to 1,459.25 ha (29%) in a southeast, south, southwest, west, and northwest direction. Chachapoyas expanded from 924.50 to 1138.05 ha (23%) in the southwest, north, northeast, and southeast directions. The study presents an analytical method integrating cloud processing, GIS, and change simulation modeling to evaluate urban growth spatio-temporal patterns and LULC changes. This approach effectively identified the main LULC changes and trends in the study area. In addition, potential urbanization areas are highlighted where there are still opportunities for developing planned and managed urban settlements.

Urban growth and Land Use/Land Cover (LULC) changes have increased in recent decades due to anthropogenic activities. This study explored past and projected future LULC changes and urban growth patterns in the Bagua Grande and Chachapoyas districts using Landsat imagery, cloud computing, and predictive models for 1990 to 2031. The analysis of satellite images was grouped into four time periods (1990–2000, 2000–2011, 2011–2021 and 2021–2031). The Google Earth Engine (GEE) cloud-based system facilitated the classification of Landsat 5 ETM (1990, 2000, and 2011) and Landsat 8 OLI (2021) images using the Random Forest (RF) model. A simulation model integrating Cellular Automata (CA) and an Artificial Neural Network (ANN) Multilayer Perceptron (MLP) in the MOLUSCE plugin of QGIS was used to forecast urban sprawl to 2031. The resulting maps showed an overall accuracy (OA) of over 92%. A decrease in forested area was observed, from 20,807.97 ha in 1990 to 14,629.44 ha in 2021 in Bagua Grande and from 7,796.08 ha to 3,598.19 ha in Chachapoyas. In contrast, urban areas experienced a significant increase, from 287.49 to 1,128.77 ha in Bagua Grande and from 185.65 to 924.50 ha in Chachapoyas between 1990 and 2021. By 2031, the urban area of Bagua Grande is expected to increase from 1,128.77 to 1,459.25 ha (29%) in a southeast, south, southwest, west, and northwest direction. Chachapoyas expanded from 924.50 to 1138.05 ha (23%) in the southwest, north, northeast, and southeast directions. The study presents an analytical method integrating cloud processing, GIS, and change simulation modeling to evaluate urban growth spatio-temporal patterns and LULC changes. This approach effectively identified the main LULC changes and trends in the study area. In addition, potential urbanization areas are highlighted where there are still opportunities for developing planned and managed urban settlements.