Fractal Geometry Analysis of Vegetation Cover Changes in Niger State Using Remote Sensing Data

Desertification and land degradation in the northern senatorial district of Niger State, North Central Nigeria, are progressively reducing vegetation cover with consequences for agricultural livelihoods and regional climate patterns. This study applies fractal geometry analysis, specifically box-counting dimension estimation, to quantify the spatial complexity and fragmentation of vegetation cover patterns using Landsat 8 satellite imagery for the years 2015, 2018, 2021, and 2023. Images are processed using the QGIS software platform with the NDVI index applied to classify vegetation presence. The binary vegetation maps generated for each year are analysed using the box-counting algorithm implemented in Python to compute fractal dimensions at spatial resolutions of 30, 90, and 270 metres. Results reveal a statistically significant decrease in fractal dimension from 1.74 in 2015 to 1.61 in 2023, indicating progressive fragmentation and simplification of vegetation patterns. The rate of fragmentation accelerates in the post-2019 period, coinciding with reduced annual rainfall documented by the Nigeria Meteorological Agency. Correlation analysis links fractal dimension values with NDVI-derived vegetation density indices across districts. The study demonstrates that fractal dimension provides a sensitive, non-parametric indicator of landscape degradation complementary to area-based coverage statistics. Keywords: fractal geometry, vegetation fragmentation, desertification, Niger State, remote sensing.