High-resolution modelling of Chennai’s 2015 extreme rainfall event shows that accounting for aerosol–cloud interactions can significantly improve rainfall and flood simulations, with potential applications in reservoir and urban flood management
CHENNAI : An Indian Institute of Technology Madras (IIT Madras)-led international study has found that atmospheric air pollution could play a critical role in improving the forecasting of rainfall and urban floods.
The study shows that explicitly accounting for atmospheric aerosols—tiny particles suspended in the air that can influence cloud formation and rainfall—can substantially improve simulations of extreme rainfall and the resulting urban flooding.
The research was carried out through a multidisciplinary collaboration between IIT Madras, GFZ Helmholtz Centre for Geosciences, Germany, Japan Aerospace Exploration Agency (JAXA), and Kathmandu University, bringing together expertise in atmospheric science, hydrology, hydraulic modelling and Earth observation.
The researchers used Chennai’s December 2015 extreme rains and flooding event as a case study. By incorporating aerosols and cloud condensation nuclei (CCN), particles around which cloud droplets form, into high-resolution atmospheric simulations, they investigated how changes in atmospheric aerosol conditions affect rainfall and, subsequently, runoff, reservoir inflows and flood inundation.
The findings were published in Natural Hazards and Earth System Sciences (https://doi.org/10.5194/nhess-26-3749-2026), a peer-reviewed, open-access international journal covering research on natural hazards, their impacts on society, environment, processes and interactions within the Earth system that drive hazard risk.
The Research Paper was co-authored by Dr. N. Nithila Devi from Department of Civil Engineering, IIT Madras and Section 4.4: Hydrology, GFZ Helmholtz Centre for Geosciences, Germany, Dr. Rakesh Teja Konduru from Earth Observation Research Center, Japan Aerospace Exploration Agency (JAXA), Dr. Kundan Lal Shrestha from Kathmandu University, Nepal, and Mr. Oscar Paul, Prof. Soumendra Nath Kuiry, and Prof. Chandan Sarangi (corresponding author) from IIT Madras.
Explaining the significance of the research, Dr. Chandan Sarangi, Department of Civil Engineering, IIT Madras, and corresponding author, said, “Urban flood forecasting is often treated primarily as a rainfall-to-runoff problem. The problem in urban flood forecasting is that rainfall spatial and time distribution is not well captured by our weather models. What happens in the atmosphere before and during rainfall can significantly influence the spatial rainfall pattern. By explicitly representing aerosol–cloud interactions, we can better capture the spatial and temporal characteristics of extreme rainfall within urban regions and consequently improve flood inundation simulations”
Prof. S. N. Kuiry, Department of Civil Engineering, IIT Madras, added, “For a megacity like Chennai, rainfall forecasting cannot be considered in isolation from water-resources management. Timing and spatial distribution of extreme rainfall directly influence runoff, reservoir inflows and flood forecasting. A coupled atmosphere–hydrology–hydraulic framework can provide a more physically informed basis for these interconnected decisions.”
The study used the Weather Research and Forecasting (WRF) model to simulate the atmospheric processes associated with the 2015 Chennai event under different aerosol conditions. The resulting rainfall fields were then coupled with HEC-HMS for hydrological modelling and HEC-RAS for hydraulic and flood-inundation modelling over Adyar basin.
The results showed that explicitly representing aerosols led to an approximately 22% improvement in simulated rainfall over the Adyar basin compared with the control simulation. This improvement propagated through the modelling chain, resulting in an approximately 50% improvement in the accuracy of simulated flood inundation compared with the control case.
KEY IMPLIFICATIONS
The findings could have important implications for water-resources management in Chennai, particularly during extreme rainfall events when decisions on reservoir storage and release are time-sensitive.
For a rapidly urbanising coastal city such as Chennai, maintaining adequate reservoir storage during the monsoon is a complex balancing act.
The IIT Madras-led study highlights the importance of obtaining more accurate urban-basin-scale rainfall forecasts, particularly for intense rainfall events.
The collaboration combined atmospheric modelling, hydrological and hydraulic simulations, and satellite-based Earth observation to examine the complete pathway from atmospheric processes to urban flooding. The research was supported by the Asia-Pacific Network for Global Change Research (APN).
The researchers emphasise that the framework is not yet an operational flood-forecasting system. Further studies are required to establish whether the improvements observed for the 2015 Chennai event can be consistently reproduced across other extreme rainfall events and in other cities.
Future work will also require better high-resolution observations of atmospheric aerosols, CCN and urban emissions, which can help improve model representation and validation. Making the computationally intensive simulations sufficiently efficient for operational forecasting will be another important challenge.
Nevertheless, the study provides a potential pathway towards more comprehensive urban flood-forecasting systems by linking atmospheric pollution processes with conventional rainfall, hydrological and hydraulic models.
The researchers note that the approach could eventually be explored in other rainfall-prone megacities, particularly those where rapid urbanisation, intense precipitation and constrained water-storage capacity combine to create complex flood-management challenges.
