The Modifiable Areal Unit Problem (MAUP) in Geographic Information Systems

(An Applied Study on Population Density)

Abstract

The study analyzes the impact of the spatial scale effect as a main dimension of the Modifiable Areal Unit Problem (MAUP) using 2022 population data of Jazan City, assessing how changing grid size affects spatial analysis and emphasizing the need to consider scale in aggregated data. Following a spatial analytical methodology, the work began with building three spatial grids (200 m, 400 m, 800 m) based on the 100 m grid, then population density was calculated in the four areal units, then scale-effect analyses were carried out through descriptive statistics and Global and Local Moran’s I to produce charts and maps. Results show that mean population density decreases gradually as grid size increases and that deviation from normal distribution declines in the 800 m grid, making the distribution appear relatively symmetrical. Maps reveal that larger spatial units change the density pattern: smaller grids give a more precise and detailed view of spatial variation, while larger grids yield a more generalized and homogeneous pattern and lose fine-scale detail. Global Moran’s I values vary with scale: the 100 m grid shows strong local spatial clustering, and the 800 m grid also shows clustering but likely from broader, less detailed groupings. Local Moran’s I scatterplots indicate that spatial autocorrelation between population density and neighboring values is affected by unit size. Local Moran’s I maps display a gradual pattern of the scale effect within MAUP. With the fine 100 m grid, the maps capture numerous high- and low-density clusters and detailed local transitions, whereas at 800 m clusters are minimal and local details almost absent, reflecting increased generalization and loss of fine-scale relationships. The findings confirm the significant influence of spatial scale on MAUP analysis and recommend multi-scale analysis and explicit consideration of MAUP effects in spatial studies.