On 16 August, 2018, the southwestern Indian state of Kerala was hit by severe floods caused by extreme rainfall during the country’s monsoon season. It was the worst flood in Kerala in more than a century.
According to the Kerala government, one-sixth of the state’s population was impacted by the floods. The freak event claimed 483 lives, displaced roughly 5.4 million people, and shut down the state’s airport, which was built on a floodplain.
India ranks as one of the world’s most flood-prone countries. With a predicted higher frequency of extreme events, and more than 12% of its geographical area prone to floods, Indian communities desperately need life-saving, assistive technology that can help prepare for these kinds of freak weather events.
Existing technology can observe and map flooding, and predict heavy rainfall, using remote sensing satellites that provide a synoptic view of large spatial regions periodically, but cannot accurately predict what areas will flood.
PhD engineering student Kesav Unnithan, from the IITB-Monash Research Academy, said that while water level changes at specific sites in India can be tracked using gauges, “flooding is not uniform from city to city, and region to region”.
“We need to devise ways to apply new technology in order to better prepare communities at risk of flooding.”
He says that groundbreaking technology, used by researchers at IITB-Monash Research Academy, has the potential to do this, and that it will help authorities in India save lives and communities at risk of extreme flooding and other natural disasters.
“Using readily-available near-real time satellite data, we were able to identify the presence of surface water across India, and have created a flood-mapping tool that can help authorities warn communities of imminent threats to life due to extreme rainfall,” he says.
"This model can measure and record two aspects of flooding. Firstly, it can track flooding across large geographical areas. Secondly, because flooding can impact various areas in varying degrees due to differences in terrain, this model can take into account the topographical impact of flooding for a given area."
He says that with further testing, this flood-mapping technology has the potential to be used internationally to warn and prepare communities of extreme weather events, such as floods caused by high rainfall, cyclones, and typhoons.
How does the technology work?
The flood-mapping technique relies on remote sensing technology called the Global Navigation Satellite System – Reflectometry (GNSS-R) that analyses GPS signals reflected off the Earth’s surface.
CYGNSS is a NASA eight-satellite constellation that was launched in December 2018, and uses GNSS-R to understand hurricane and cyclone behaviour.
These signals, which have near-global coverage and are free, were later found to be able to sense the presence of surface water, and so were suitable for flood mapping.
Currently, this method is exploratory, as it focuses on improving the poor spatial resolution of the CYGNSS data set in mapping flood extents.
But still, this method is a useful way to understand, and help the scientific community improve accuracy of potential floods.
“Furthermore, with the future availability of improved CYGNSS-like GNSS-R datasets, our technique can potentially be used for operational purposes, and can provide affected people with information to make decisions that are more effective,” he says.
The climate is changing
A flood occurs when water inundates typically dry land. Most floods take hours or days to develop, which gives residents sufficient time to evacuate. Others generate quickly and with little warning.
Floods increasingly affect communities in developing countries, especially on the Indian subcontinent. This is mainly due to extreme precipitation events, which have increased three-fold since 1950 due in part to climate change, combined with poor infrastructure.
Severe and prolonged rainfall events, with higher frequencies, can inundate larger areas, causing high fatalities and extensive damage to property. Downstream effects due to flooding include loss in agricultural products and commercial production and income – all integral to the Indian economy.
Residents of flooded areas are left without power and clean drinking water, leading to outbreaks of deadly water-borne diseases such as hepatitis A and cholera.
Efforts are being made across the world to address flood risks to vulnerable communities, including making more buildings flood-resistant, and developing warning systems to evacuate communities and deploy assistance expeditiously.
Saving lives one map at a time
“Our current research empowers people to make more informed decisions in the face of natural disasters,” Kesav Unnithan said.
“Once we have access to future predicted rainfall values from climate models, hydrological models can be used to convert them into runoff values.
“The inundation model then makes assumptions on river velocity and geometry, to have quick but detailed analysis for a vast region of interest, without compromising on the accuracy of predicted flood extents.
“By employing a simple yet effective routing technique to address the transport of discharge and converting this discharge to inundation, the final output can be visualised as a probability map of flooding on a given date.
“Because this entire process is made time-efficient, these probability maps can inform people beforehand of impending floods,” he said.
“Once we’ve trialled the inundation model across different topographic conditions and predicted rainfall values, we can then have a faster, more efficiently predicted probability of flooding, mapped with a future rainfall event for any region in the world.”
This research project is supervised by Professor Basudev Biswal in the Department of Civil Engineering at IIT Bombay, and by Professor Christoph Rüdiger in the Department of Civil Engineering at Monash University. It's funded by CSIRO and externally supervised by Dr Mahesh Prakash, Senior Principal Research Scientist, from CSIRO Data61.
To receive a fortnightly email wrap up of stories from Lens.