How geomatics can be used to accurately assess urban flash flood risks

Rapid urbanisation at the cost of waterbodies, unplanned settlements and changing weather patterns are increasing the risk of flash floods and similar problems in Chennai, Delhi and Ahmedabad and other Indian cities.

Hyderabad, Chennai, Mumbai, Surat, Kolkata and Srinagar have already witnessed flash floods in recent decades and the growing threat calls for urgent sustainable solutions.

The challenges of flash floods extended globally to arid regions like Al-Ain City, United Arab Emirates (UAE), Saudi Arabia, Yemen, Oman and Egypt, where urban growth and climatic extremes increase vulnerabilities. Addressing these challenges is vital for ensuring the safety and sustainability of urban life.

Al-Ain, located in the arid region of UAE, experienced unexpected flash floods in 2024 due to severe storms, underscoring the need for a comprehensive flood-risk assessment. “It is important in the first instant to understand what a ‘flood’ is. Hydrologically, flood is understood as a flow concept, and what overflows over and above the riverbed and flood plains is considered as flood. But in the ordinary use everything is called flood including water stagnations and inundations,” said S Janakrajan, president, SaciWATERs.

Many lakes in India are encroached for building urban infrastructure, housing, among other things, the expert noted. “When these areas attract huge inundations / water stagnation, we call it flood. Therefore, urban floods / inundations are manmade and occurs mainly because of distortions created in the given hydraulics and hydrology of a city or a town."

What’s more, all cities and towns should carefully engage in not just land use management but evolve an integrated approach in which land use should be integrated with urban water management, he added.

A study published in Science Direct journal assessed flood vulnerability in Al-Ain, UAE by integrating geographic information systems (GIS), remote sensing, analytical hierarchy process and multi-criteria decision analysis (MCDA). Led by Mona S Ramadan, a researcher from the Geography and Urban Sustainability Department at UAE University, the study identified high-risk flood areas and evaluates factors like population density, impervious surfaces, elevation and rainfall intensity, using spatial modeling to guide flood management strategies.  

The researchers also integrated spatial datasets, such as the Advanced Spaceborne Thermal Emission and Reflection Radiometer Digital Elevation Model (ASTER DEM) acquired from NASA Earth Data in 2021 with 30 metre resolution, provided essential topographical information, including elevation, slope, flow direction and flow accumulation.

Additionally, rainfall data was obtained in 2024 with a resolution of 0.1 degree from the Global Precipitation Measurement and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks satellite missions, which are primarily used to monitor rainfall, droughts, floods and soil moisture.

This data was processed in ArcGIS software to quantify rainfall intensity and accumulation. These parameters were critical for understanding water movement patterns during rainfall events, helping identify low-lying areas prone to flooding. Sentinel-2 imagery of 2024, with a 10-metre resolution, was also used to classify land cover, monitor urban growth and calculate vegetation indices to assess flood risk.

Preprocessing steps, such as image quality correction and cloud masking, enhanced data accuracy.

The ASTER DEM was also used to generate slope maps and flow direction models, further helping to identify areas vulnerable to flooding. In the study, the author used AHP, a decision-making method that evaluates different options against multiple criteria, to identify the best alternative. AHP assigns weights to factors like rainfall intensity, elevation, slope, land cover and population density.

These weights were determined through a pairwise comparison matrix, ensuring a consistent evaluation of each factor’s importance. The parameters were then grouped into three vulnerability levels: Low, moderate and high.

After completing all the processes, a flood vulnerability map was created using ArcGIS software that showed areas with high, moderate and low risk, considering all the factors. High-risk areas, often with dense urbanisation, low elevation and impermeable surfaces, were identified as the most vulnerable. These areas needed urgent attention for infrastructure development and flood mitigation.

The high population density further increased the potential impacts of flooding, such as human casualties and infrastructure damage, the author added.

In the era of climate change, urban flash floods are becoming a bigger problem globally, especially in dry areas like Al-Ain. Geospatial tools such as GIS, remote sensing, AHP and multi-criteria decision analysis will help urban planners to better understand flood risks and create more effective mitigation strategies.

By utilising these technologies, cities can move toward a future that is not only flood-resilient but also more sustainable and adaptive to climate. In India, Murugesan Bagyaraj used MCDA and AHP to assess the flash flood risk in Chennai in 2023. The results and methodology closely align with the Al-Ain study, particularly in highlighting the effects of urbanisation and inadequate drainage systems on flash flood susceptibility. 

Rohith AN, assistant professor, Indian Institute of Technology Delhi, pointed out that despite examples like the Chennai flash flood assessment by Murugesan Bagyaraj, a significant challenge Indian researchers face in flood vulnerability assessment is the absence of high-resolution data, including land use, soil and elevation, as well as limited hydro-meteorological data such as stream flow, rainfall and temperature.

Additionally, India’s restrictive data-sharing policies often require researchers to purchase critical datasets, which are not always available, even for research purposes. This limitation hinders the ability to test methodologies locally, frequently forcing researchers to depend on data from US watersheds. While initiatives like India-Water Resource Information System have been introduced to improve data accessibility, they have yet to fully address the problem.