The mapping exercise went beyond the official record. According to the Delhi Parks and Gardens Society and the Wetland Authority of Delhi, 1,009 waterbodies were officially documented within the territory. The satellite-based inventory identified 1,167 waterbodies, of which 793 corresponded to locations in the official inventory and 374 were identified only from imagery, meaning 158 sites beyond the official count. The authors caution that some newly mapped features may reflect seasonal water occurrence, land-use change, differences in mapping criteria, or updates to administrative records. Still, the expanded inventory is intended to prevent restorable sites from being excluded from planning.

Accuracy was tested in the field. Approximately 10 percent of the mapped sites, representing 115 waterbodies, were selected through stratified sampling across Delhi’s eleven administrative districts and visited to confirm their physical presence and general condition as active, seasonal, degraded, or encroached. Site locations were verified with a Garmin GPS receiver accurate to roughly three meters under open-sky conditions. The comparison between mapped inventory and field observations yielded an overall accuracy of 92 percent and a Kappa coefficient of 0.88, indicating strong agreement.

The analytical core of the study is a newly formulated Terrain Connectivity Index, or TCI, designed to quantify relative terrain-supported restoration suitability for every mapped waterbody. The index combines four parameters: the elevation difference between each waterbody and the nearest relevant river corridor, the slope gradient, the upstream catchment contribution derived from flow accumulation, and a topographic barrier condition reflecting obstruction from dense built-up areas, transport corridors, and modified drainage channels. Each parameter was normalized to a scale from zero to one using min-max normalization, and the weighted combination was expressed as TCI equal to 0.4 times the normalized elevation difference, plus 0.3 times the normalized slope, plus 0.2 times the normalized upstream catchment contribution, plus 0.1 times the normalized barrier condition.

The weighting scheme deserves scrutiny. The authors assigned the weights a priori based on the conceptual role of each parameter in terrain-supported runoff movement, giving elevation difference the greatest weight because terrain gradients influence the potential movement of runoff toward low-lying waterbodies. No formal statistical calibration, pairwise comparison, or optimization procedure was used, so the weights should be read as study-defined relative importance factors rather than empirically calibrated coefficients. A sensitivity analysis varying the weights within a plus-or-minus 10 percent range found that the overall spatial prioritization pattern remained largely stable under moderate variation, and a qualitative comparison with post-monsoon water persistence patterns showed general spatial agreement.

Across all 1,167 sites, TCI values ranged from 0.15 to 0.87, revealing substantial spatial variation. The index was classified into three relative suitability bands, with values up to 0.30 deemed low suitability, 0.31 to 0.60 moderate, and above 0.60 comparatively favorable. For reporting, the low and moderate classes were combined as hydrologically constrained, while the highest class was designated terrain-supported. The authors emphasize that this two-group labeling does not mean low- and moderate-scoring sites are unsuitable for restoration; it indicates only that their terrain-supported conditions are comparatively less favorable within this dataset. Terrain-supported sites were concentrated in low-lying areas associated with the Yamuna floodplain and parts of western Delhi, while constrained sites were more widely distributed across elevated and heavily urbanized parts of the city.

District-level results showed pronounced contrasts. East Delhi had the highest proportion of waterbodies classified as terrain-supported at 65 percent, followed by Southeast Delhi at 35 percent, Shahdara at 32 percent, and Southwest Delhi at 26 percent. South Delhi and New Delhi were largely constrained by elevated topography, fragmented drainage, and deeper groundwater tables. The authors also highlight the importance of reading percentages alongside absolute numbers: Southwest Delhi contains the largest number of mapped waterbodies at 295, but only 26 percent fall in the favorable category, so the district may still represent an important planning area because of the sheer number of sites available for intervention.

Groundwater depth added a second, complementary layer to the assessment. Records from the Central Ground Water Board, expressed in meters below ground level, were classified into four categories: shallow at 0 to 10 meters, moderate at more than 10 to 20 meters, deep at more than 20 to 40 meters, and very deep at more than 40 meters. District averages varied widely. Northwest Delhi showed the shallowest average depth at approximately 3.7 meters with 177 mapped waterbodies, while Central Delhi averaged about 4.8 meters with 59 waterbodies. Southwest Delhi averaged roughly 9.9 to 10 meters across 295 sites, East Delhi about 10.5 meters, North Delhi approximately 10 meters, Southeast Delhi 9.8 meters, Northeast Delhi 16.6 meters, and Shahdara 14.4 meters. New Delhi averaged approximately 21.6 meters across 96 waterbodies, and South Delhi presented the deepest conditions at roughly 44 meters across 99 waterbodies.

These groundwater patterns shaped the suggested restoration contexts, though the authors are careful about interpretation. Groundwater depth was treated as a supporting spatial indicator, not as a direct measure of hydraulic connectivity or recharge potential. Shallow groundwater was considered comparatively favorable for controlled recharge interventions where local hydrogeological conditions permit, whereas deeper groundwater was interpreted as an increasing constraint on groundwater-supported restoration, pointing instead toward stormwater harvesting, surface-water retention, treated wastewater reuse, or re-routed drainage. Long-term district-level observations from 2000 to 2024 showed that the Yamuna floodplain and adjoining low-lying areas generally maintained comparatively shallow groundwater, while South Delhi and parts of New Delhi exhibited more persistent decline. The authors stress that the association between shallow groundwater and low-lying terrain is spatial and descriptive and should not be read as evidence that shallow groundwater guarantees successful recharge.

The study’s limitations are stated plainly. The 30-meter Cartosat DSM may not resolve micro-topographic features, small embankments, culverts, drains, or building-scale flow controls that influence urban runoff. The June 2023 Sentinel-2 inventory represents a single observation period, and seasonal variation can affect the apparent number and extent of waterbodies. The barrier parameter represents a relative assessment of drainage continuity rather than a hydraulically calibrated obstruction coefficient, and the TCI was not calibrated against measured discharge, flow velocity, hydraulic connectivity, or restoration outcomes. Ground validation is stronger for the mapped inventory, based on the 115 field-verified sites, than for the TCI, whose validation remains qualitative. The index is a comparative screening and prioritization framework, not a hydraulic prediction model, and it does not directly quantify discharge, flow velocity, inundation depth, or groundwater recharge.

Even so, the framework offers something Delhi’s waterbody management has lacked: a city-scale, spatially explicit way to move beyond static inventories. Existing records confirm where a waterbody is or was, but they do not indicate whether the surrounding landscape can physically support hydrological restoration. By coupling terrain topology, drainage connectivity, urban barriers, and localized groundwater depth in a single reproducible index, the study enables planners to differentiate self-sustaining, terrain-favored waterbodies from heavily constrained sites that will require engineered stormwater re-routing or artificial recharge. The authors frame the district-level intervention suggestions as planning-level options only, noting that final selection requires site-specific assessment of drainage infrastructure, water quality, land availability and ownership, aquifer characteristics, and calibrated hydrological or hydraulic modeling. For a megacity where decades of uncoordinated urbanization have decoupled ponds, johads, and floodplain depressions from their catchments, that distinction between naturally viable and engineered restoration could determine which of Delhi’s 1,167 waterbodies are worth saving first.

Subject of Research: Terrain-based prioritization of urban waterbody restoration in Delhi using topography, drainage connectivity, and groundwater conditions

Article Title: Terrain based prioritization of urban waterbody restoration using topography drainage connectivity and groundwater conditions in Delhi

Article References: Jha, M. K., Sahu, V., & Kainthola, A. (2026). Terrain based prioritization of urban waterbody restoration using topography drainage connectivity and groundwater conditions in Delhi. Discover Geoscience, 4(1), Article 417. https://doi.org/10.1007/s44288-026-00797-8

Image Credits: AI Generated

DOI: 10.1007/s44288-026-00797-8

Keywords: Delhi, urban waterbodies, Terrain Connectivity Index, restoration prioritization, GIS mapping, Sentinel-2, Cartosat DSM, groundwater depth, drainage connectivity, Yamuna floodplain, hydrologically constrained sites, urban hydrology