§ 01
PROBLEM
PROBLEM
What it had to solve.
Urban growth indicators from census data lag by 5–10 years in developing cities. Planners need something more responsive — and globally consistent.
Urban Sprawl — satellites watching cities grow.
Multi-year satellite analysis quantifying urban sprawl trajectories at the city scale. Segmentation fused with change detection to produce longitudinal sprawl indices.
Years analyzed
15
2005–2020
Cities
3
Pakistan
Pipeline stages
4
Seg · Change · Index · Viz
Correlation
r=0.89
vs census growth
§ 02
APPROACH
APPROACH
How it works.
Landsat imagery from 2005–2020 is segmented frame-by-frame; change detection between adjacent years produces "growth pixels"; a sprawl index aggregates to the municipal level.
§ 03
RESULTS
RESULTS
What it delivered.
Derived sprawl indices correlated with census growth (r = 0.89) while providing annual resolution — five years ahead of the next census.