Water obeys terrain, and GIS hydrology formalises that obedience. From a terrain model, flow-direction and flow-accumulation algorithms derive where water travels and concentrates; from any outlet, the contributing watershed can be delineated; stream networks emerge from accumulation thresholds. These primitives feed everything from drainage design and erosion modelling to the flood studies that protect cities.
The core derivation chain
The classic chain runs: condition the DEM (fill spurious pits, enforce known drainage), compute flow direction per cell, accumulate flow downstream, extract streams above a contributing-area threshold, and delineate watersheds to any pour point. Each step has algorithmic choices (D8 versus multi-directional flow) with real consequences in flat and complex terrain.
Outputs — basins, stream orders, flow paths, wetness indices — become the skeleton of hydrological and environmental models.
Why terrain preparation dominates
Raw DEMs sabotage hydrology: road embankments dam digital rivers where culverts exist unseen; bridges block valleys; noise creates false pits. Hydro-enforcement — cutting drainage through structures, flattening water bodies, enforcing monotonic streams — is the specialised editing that makes terrain hydrologically honest.
This is why flood programmes specify hydro-enforced or hydro-flattened DEMs explicitly, and why LiDAR terrain plus enforcement has become the flood-mapping standard.
Toward flood and design applications
GIS supplies hydraulic models (HEC-RAS and kin) their geometry — terrain, cross-sections, roughness from land cover — and maps their results back as inundation extents and depth grids joined to exposure: people, buildings, infrastructure at risk by scenario.
The same chain serves design: catchment areas for culvert sizing, runoff estimation with land-cover coefficients, watershed prioritisation for treatment programmes.