Urban-Flood Risk Mapping of Chennai's Adyar Sub-Basin Using AHP-Weighted GIS Overlay (QGIS)

This M.Plan thesis builds a ward-level urban-flood risk map for the Adyar sub-basin in Chennai and turns it into land-use and development-control recommendations. Chennai's floods of December 2015 showed that heavy rain alone does not explain the damage: the city had built over tanks, marshes and floodplains, narrowed channels and paved the catchment, so water had nowhere to go and people had nowhere to escape.

The thesis treats flood risk as the product of hazard and vulnerability. Hazard layers — elevation and slope from CartoDEM or SRTM, drainage density, distance to the river and major channels, IMD gridded rainfall, land use/land cover classified from Sentinel-2 imagery and soil hydrological group — are reclassified and weighted using Saaty's Analytic Hierarchy Process (AHP), with a consistency check on the pairwise matrix. Vulnerability layers are built from Census 2011 ward data (population density, share of slum population, households without pucca housing) and a map of critical facilities such as hospitals, schools and electrical substations.

Everything is done in QGIS with open data from Bhuvan, USGS and the Census. The hazard map is validated against publicly released inundation maps of the 2015 event, and the final risk zones are used to propose a flood-sensitive zoning overlay, buffer rules along channels, restoration of water bodies and priority wards for drainage investment, framed within NDMA's urban-flooding guidelines and URDPFI 2015.

Abstract

Urban flooding in Indian cities is increasingly a planning failure rather than a purely hydrological one. This thesis develops a replicable GIS-based method to identify flood-risk zones at ward level in the Adyar sub-basin of Chennai by combining hazard and vulnerability indicators through an AHP-weighted overlay in QGIS. The hazard component is validated against mapped inundation from the 2015 floods. The output is a risk-zonation map and a set of land-use, development-control and infrastructure recommendations aimed at the city's master-planning process.

Introduction

The Adyar river drains a large part of south and west Chennai, including fast-growing peri-urban areas around the airport and the IT corridor. Its catchment once held a dense network of irrigation tanks and marshes that stored monsoon water. Rapid urbanisation converted many of these into layouts, roads and institutional campuses. When the north-east monsoon brings intense rainfall, runoff reaches the river faster and in greater volume than the channel and the remaining storage can handle.

Literature gap

GIS and multi-criteria methods for flood-hazard mapping are well established, and several studies have mapped Chennai's 2015 inundation. Fewer studies take the next planning step: combining hazard with social vulnerability at the ward level and translating the result into development-control language that a master plan could adopt. Many hazard studies also stop without validation. This thesis addresses both gaps with a validation step and a planning-proposals chapter.

Existing approach vs proposed approach

• Existing: flood response is largely engineering-led — desilting, new drains, relief works — and land-use plans rarely carry a flood-risk overlay. Development permissions are granted plot by plot without a basin-level view.
• Proposed: a transparent, reproducible risk map that planners can update, with weights that are documented and tested, and zoning categories that link each risk class to permitted uses, plinth levels, ground-coverage limits and mandatory on-site retention.

Feasibility

• Data: all core datasets are public — CartoDEM from Bhuvan, SRTM and Landsat from USGS EarthExplorer, Sentinel-2 from Copernicus, gridded rainfall from IMD, and ward tables from the Census of India.
• Technical: QGIS with GRASS and SAGA tools handles hydrological processing, reclassification and weighted overlay on an ordinary laptop.
• Institutional: recommendations are framed in terms already used by Indian planning bodies (URDPFI land-use categories, NDMA guidelines), so they can be discussed with a development authority.

Large parts of the Adyar sub-basin in Chennai flood repeatedly during intense north-east monsoon rainfall, with severe loss of life, property and livelihoods in the 2015 event. Flood exposure is not evenly distributed: low-lying areas that were once tanks or floodplains now carry dense housing, including informal settlements with the least capacity to cope and recover. Yet the city's statutory land-use planning does not carry a systematic, ward-level flood-risk layer, and development permissions continue in areas that function hydrologically as storage.

The problem this thesis addresses is: which wards and land parcels in the Adyar sub-basin carry the highest combined flood hazard and social vulnerability, and what land-use, development-control and infrastructure measures should the master plan adopt for each risk class? The answer must come from a documented, reproducible method with validated results, not from a one-off map.

In scope

• The Adyar sub-basin within the Chennai Metropolitan Area, analysed at a 30 m raster resolution and summarised by ward.
• Pluvial and fluvial flooding driven by monsoon rainfall; hazard by multi-criteria overlay, not hydraulic modelling.
• Vulnerability from Census 2011 (the latest available ward-level census) plus a critical-facilities inventory.
• Planning recommendations: zoning overlay, development-control provisions, water-body restoration and priority drainage works.

Out of scope

• Hydrodynamic modelling (HEC-RAS or similar) and detailed drainage design.
• Coastal storm-surge and sea-level-rise flooding.
• Cost–benefit analysis of individual engineering works.
• Parcel-level legal review of encroachments.

The thesis uses a quantitative spatial research design with a planning-proposals stage. It runs over about 24 weeks with a review at the end of each stage.

Stakeholder input: short semi-structured interviews with residents' welfare associations and, where possible, an officer of the local body or development authority, to test whether the proposals are practical. Interview notes are anonymised.