| Abstrakt: |
Comprehending the interactions between humans and their environment necessitates modeling human–environment interactions. This study employs time-series satellite imagery from Landsat Thematic Mapper (1995 and 2009) and Landsat 8 Operational Land Imager (2022) to examine land use/land cover (LULC) change in Dar es Salaam, Tanzania, after image pre-processing with the Google Earth engine code editor, while random forest machine learning in R classified LULC. Geographically weighted regression (GWR) correlates LULC changes to socioeconomic factors spatially. Analysis reveals a dynamic LULC transformation between 1995 and 2022, with a 14.9% increase in built-up areas and a 14.6% decline in bushland. Out of the total LULC, 65.8% experienced gains and losses, while 34.2% remained stable. The GWR model, surpassing the ordinary least squares (OLS) model, achieves an R2 value of 0.73, indicating a strong relationship between LULC changes and socioeconomic factors, explaining 73% of the variation. The influences of these factors exhibit variations across different LULC change types. Population density and proximity to the city center significantly contribute to LULC changes, while the impacts of gross domestic product and distance to roads are comparatively less significant. Poverty does not drive LULC changes significantly. The findings indicate that urbanization and urban sprawl, influenced by population density and distance from the city center, significantly impact land use and cover changes. Effective urban planning strategies should be prioritized to address this, considering factors such as population density and distance from the city center to mitigate the considerable effects on land use and cover changes in the study area. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink