Application of GIS & Remote Sensing in Water Resources Engineering

Introduction: Why Geospatial Technology Matters for Water

Water resources engineering is inherently spatial. Rainfall falls over catchments. Rivers flow through valleys. Groundwater moves through aquifers defined by geology and topography. Irrigation systems serve farms across thousands of hectares. Every question a water resources engineer asks — How much runoff will this catchment generate? Where will the floodwaters go? Is this dam still stable? — has a spatial answer.

For most of the twentieth century, answering these questions required extensive, expensive field surveys. Topographic maps were hand-digitised. Rainfall data came from sparse gauge networks. Land use was mapped from aerial photographs interpreted by trained analysts. The cost and time involved meant that many design studies worked with incomplete data, and many infrastructure projects in data-scarce regions were designed on professional judgment rather than systematic spatial analysis.

GIS and remote sensing have changed this calculus dramatically. Today, a water resources engineer with a laptop and an internet connection can access global datasets of topography, rainfall, land use, soil type, vegetation, soil moisture, river discharge, and groundwater storage — often at no cost. Cloud computing platforms like Google Earth Engine allow analysis of petabytes of satellite imagery without downloading a single file. These capabilities have democratised sophisticated hydrological analysis and are transforming practice, particularly in the developing world where ground-based data networks remain sparse.

1. Catchment Delineation and Morphometric Analysis

The catchment — the area of land draining to a common outlet — is the fundamental unit of hydrological analysis. Delineating catchments accurately is a prerequisite for almost every subsequent analysis: runoff estimation, flood frequency analysis, sediment yield estimation, and water balance modelling.

DEM-based delineation

Modern catchment delineation uses digital elevation models (DEMs) to simulate flow direction and accumulation across a landscape. The process involves filling sinks, computing flow direction grids, accumulating flow, and snapping pour points to the stream network — all operations that GIS software handles automatically from raw elevation data.

Key DEM sources for this work include:

• SRTM (Shuttle Radar Topography Mission) — 30m and 90m resolution, global coverage, freely available
• ALOS PALSAR — 12.5m resolution, better performance in forested terrain
• Copernicus DEM (GLO-30) — 30m, currently among the most accurate freely available global DEMs
• FABDEM — forest and buildings removed DEM, derived from Copernicus DEM; particularly useful for hydraulic modelling
• TanDEM-X — 12m commercial product, superior accuracy for professional projects

Morphometric analysis

Once delineated, a catchment can be characterised by a suite of morphometric parameters — drainage area, main channel length, basin slope, stream order, drainage density, shape factor, and time of concentration. These parameters drive hydrological model parameterisation and can be extracted systematically using GIS tools such as the TauDEM toolset in QGIS or the Spatial Analyst toolbox in ArcGIS.

2. Satellite-Derived Rainfall and Hydrometeorological Data

Rainfall gauge networks in many developing countries are sparse and poorly maintained. Engineers designing flood protection infrastructure or undertaking water balance studies often have access to only a few gauges — or none at all — within their study area. Satellite rainfall products have partially filled this gap, though their limitations in complex terrain must be understood.

Key satellite precipitation products

Bias correction of satellite rainfall against available gauge records is a critical step before use in calibrated hydrological models. Methods range from simple scaling (multiplicative or additive) to more sophisticated quantile mapping approaches. The choice of method depends on data availability and project requirements.

3. Land Use / Land Cover Mapping

Land use determines how rainfall is partitioned between infiltration, evapotranspiration, and runoff. In the SCS Curve Number method — the most widely applied runoff estimation technique globally — land use and soil type are the primary inputs. In distributed hydrological models, spatially variable land use grids control infiltration, evaporation, and routing parameters across the catchment.

Classification approaches

Remote sensing provides two main avenues for land use mapping:

• Pre-classified global products: ESA WorldCover (10m), GlobCover, MODIS MCD12Q1, and Copernicus Global Land Service provide ready-to-use classifications requiring only clipping to the study area. These are appropriate for rapid assessments and feasibility studies.
• Custom supervised classification: Using Sentinel-2, Landsat, or PlanetScope imagery, engineers can train classifiers (Random Forest, Support Vector Machine) on locally representative training samples to produce site-specific maps with higher accuracy in complex landscapes.

Change detection — comparing land use maps from different dates — is particularly valuable for assessing the hydrological impact of deforestation, urbanisation, or agricultural expansion on catchment behaviour over time.

4. Flood Mapping and Inundation Modelling

Flood mapping is one of the most impactful applications of GIS and remote sensing in water resources engineering. It informs floodplain zoning, infrastructure siting, flood risk assessment, evacuation planning, and post-flood damage assessment.

Workflow: satellite to flood map

A typical flood mapping workflow for a dam spillway design project proceeds as follows: (1) delineate the downstream floodplain using a DEM; (2) extract the cross-section geometry from the DEM for 1D HEC-RAS modelling, or use the DEM directly in a 2D model; (3) derive design floods using satellite rainfall and a hydrological model; (4) route the design flood through the hydraulic model; (5) export the inundation polygon to GIS for overlay with infrastructure, population, and land use data; and (6) produce flood hazard and risk maps for design and planning purposes.

5. Dam Site Assessment and Reservoir Planning

GIS and remote sensing support every phase of dam project development. At the reconnaissance stage, spatial analysis identifies candidate dam sites that maximise reservoir capacity relative to dam height, using DEM-derived hypsometric curves. Site screening considers geology (from geologic maps and satellite-based lithological indices), seismicity, land ownership, and upstream catchment characteristics — all of which can be assessed spatially before a single field visit.

During feasibility and detailed design, satellite imagery provides baseline land use for reservoir inundation impact assessment and resettlement planning. NDVI-based vegetation maps identify productive agricultural land within the planned reservoir area. Post-construction, SAR and optical satellite imagery monitor reservoir storage levels, sedimentation, and downstream geomorphic change.

6. Groundwater Exploration and Management

Remote sensing contributes to groundwater investigations in several ways. Structural geology interpretation from multispectral and radar imagery identifies lineaments — linear features associated with faults and fracture zones — that correlate with productive aquifers in hard-rock terrain. This technique is widely used in rural water supply programmes across sub-Saharan Africa to target borehole siting.

Gravity data from the GRACE and GRACE-FO satellite missions provides estimates of groundwater storage anomalies at regional scale, enabling tracking of seasonal and multi-year groundwater depletion trends — information of direct relevance to aquifer management and groundwater-dependent infrastructure design.

7. Irrigation System Planning and Performance Monitoring

Large-scale irrigation schemes generate distinctive spectral signatures detectable from satellite. NDVI time series from Sentinel-2 or Landsat tracks seasonal crop growth patterns and identifies non-performing or under-irrigated areas within a scheme. Evapotranspiration models driven by satellite surface temperature data (SEBS, METRIC, SSEBop) provide spatially distributed estimates of crop water demand, enabling irrigation scheduling and water allocation decisions that improve water use efficiency.

Canal condition monitoring using satellite imagery — identifying breaches, siltation, and encroachment — is an emerging application that reduces the cost and hazard of manual inspection, particularly on extensive systems.

8. Climate Change Impact Assessment

Regional climate models (RCMs) from the CORDEX programme provide downscaled projections of temperature and precipitation under various emission scenarios, typically at 25–50 km resolution. GIS tools enable the spatial analysis of these projections: mapping areas where design rainfall is expected to increase or decrease, identifying catchments at elevated flood or drought risk, and quantifying the change in seasonal runoff volumes that determines reservoir yield reliability under future climate.

Bias-corrected RCM outputs, fed into hydrological models calibrated on historical satellite and gauge data, provide the basis for climate-resilient infrastructure design — an increasingly standard requirement for projects funded by international development finance institutions.

Key Tools for the Water Resources GIS Practitioner

The practitioner needs a working toolkit. The following are the core platforms and extensions most relevant to water resources GIS work:

• QGIS — Open-source GIS with GRASS, SAGA, and WhiteboxTools integration for hydrological processing
• ArcGIS Pro — Commercial platform with Arc Hydro and HEC-GeoRAS extensions for water resources workflows
• Google Earth Engine — Cloud platform for large-scale satellite data analysis; Python and JavaScript APIs
• HEC-HMS — US Army Corps hydrological modelling; GIS integration via HEC-GeoHMS
• HEC-RAS — 1D/2D hydraulic modelling; terrain mapping via RAS Mapper
• SWAT+ — Distributed watershed model with ArcSWAT and QSWAT interfaces
• MIKE HYDRO — DHI's integrated surface-groundwater platform with GIS interface
• R (hydroTSM, rgdal, sf) / Python (geopandas, rasterio, pysheds) — For custom scripted workflows

Challenges and Limitations

Geospatial tools are powerful but not infallible. Common pitfalls in water resources GIS applications include: over-reliance on coarse global DEMs in flat terrain; uncritical use of satellite rainfall products without bias correction; failure to validate land use classifications against ground truth; and misinterpretation of model outputs as ground truth rather than as probabilistic estimates requiring engineering judgment.

The credibility of a GIS-informed design depends entirely on the quality of the underlying data and the competence of the analyst. Remote sensing literacy — understanding what satellites actually measure, how algorithms translate raw radiance into derived products, and where those algorithms break down — is a prerequisite for responsible application.

Conclusion

GIS and remote sensing have become indispensable tools in water resources engineering practice. From the desk study that characterises an ungauged catchment to the post-flood satellite analysis that quantifies inundation extent, geospatial technology is now embedded in every phase of the project cycle. Engineers who invest in building genuine geospatial competence — not just the ability to press buttons in a GIS, but a deep understanding of the data, the models, and their limitations — are equipped to deliver more credible, more efficient, and more resilient water infrastructure.