- In response to recurring extreme weather events, community-driven efforts in Kerala are establishing decentralised flood forecasting and early warning systems.
- This approach fosters efficient communication between panchayats, civil authorities, and vulnerable residents.
- Partnering with scientists for mentorship and guidance, real-time rainfall observations are shared via a messaging app and forecasts generated at a 70% accuracy rate for the entire state.
Over 100 days since the devastating landslides in Wayanad district, Kerala, which killed over 400 people, the question arises: are hyper-local weather predictions the key to mitigating the loss and damage from climate change-induced landslides and other extreme weather events?
The question is particularly relevant considering the Wayanad district authorities allegedly failed to heed warnings about the impending disaster, which were based on forecasts using hyper-local measurements of the heavy rainfall in the days leading up to the event.
While accurately predicting landslides may not be possible, climate change experts in Kerala say that early warnings can be sent out to the residents of the high rainfall regions which are landslide-prone by analysing micro-rainfall data. “Monitoring cloud movement patterns and analysing wind conditions — such as speed and direction — can lead to more precise and useful landslide forecasts, ultimately saving lives,” says Abhilash S., associate professor at the Department of Atmospheric Sciences at Cochin University of Science and Technology (CUSAT).
When did the experiment start?
After the 2019 landslides in Wayanad’s Puthumala village in Meppadi panchayat, which claimed 17 lives, the Hume Centre for Ecology and Wildlife Biology, a non-profit dedicated to protecting Wayanad’s ecosystem, established 200 rain gauges across the hill district. This initiative, carried out in collaboration with local farmers and environmental groups, aimed to collect micro-rainfall data and help prepare the community for potential rain-related disasters.
The effort proved successful during a severe weather event in 2020, when 55 rain gauges recorded 1,000 mm of rainfall in a short period near the landslide-prone Mundakkai village. Although many homes were destroyed, no lives were lost, due to the timely evacuation of hundreds of residents to safer areas.
In the aftermath of the 2024 landslides, Wayanad district authorities are facing criticism for disregarding early warnings and failing to effectively communicate the threat to both residents and local officials. It is alleged that the administration dismissed these warnings, citing their origin from unofficial sources. Yet, with repeated extreme weather events highlighting the vulnerability of the region, there is growing momentum to reassess the current approach to disaster forecasting. There is growing recognition that developing a more community-centred forecasting model is long overdue.
In an interview with Mongabay India, Kerala’s Revenue Minister K. Rajan praised the evolving weather forecasting models, noting that “they will play a key role in strengthening the state’s resilience against future natural disasters”. While acknowledging that the current weather forecasting system “needs improvement”, the minister reveals that Kerala has requested additional observational tools from the central government, including more radar systems, to enhance both accuracy and efficiency.
Decentralised and community focused
Various voluntary initiatives have been launched across the state to establish decentralised, community-driven flood forecasting and early warning systems. These systems aim to support residents in vulnerable valleys and hills, as well as those living in downstream villages along flood-prone rivers. Additionally, they are designed to help fishers in the south, who face significant risks when venturing into turbulent seas during challenging weather conditions.
“Not only Kerala, but the nation needs laws, policies, and guidelines to prompt administrations to take these warnings seriously and act on them. Real-time decentralised climate predictions can only be possible with public participation. The local bodies and grassroots-level governmental agencies can join popular movements in this regard,” says C.K. Vishnudas, founder of Hume Centre for Ecology and Wildlife Biology.
The Hume Centre has divided Wayanad into multiple grids, each measuring 25 square kilometres, and has installed at least one rain gauge in each grid. This approach allows access to localised data that is often more useful than the broader data provided by the India Meteorological Department (IMD), he explains.
Focused on river basins
On September 28, the Community-Sourced Impact-based Flood Forecast and Early Warning System (CoS-it-FloWS) was launched in the Periyar and Chalakudy river basins. Both rivers have experienced multiple floods since the major flooding event that struck Kerala in 2018. The initiative was started by Equinoct, a Kochi-based provider of community-sourced modeling solutions, in collaboration with the voluntary organisation Community Resource Centre (CRC) in Puthenvelikkara village, Ernakulam district.
“The newly launched climate information system will provide data collected daily from 100 rain gauges operated with community participation across the Periyar and Chalakudy river basins. This data will be available to the Ernakulam District Emergency Operation Centre (EOC) for disaster planning,” C. Jayaraman of Equinoct points out.
Due to repeated floods in the region, the local community has started to take proactive measures. “This collective initiative greatly enhances flood disaster preparedness, mitigation, rescue, and rehabilitation in the two crucial river basins. It also encourages other vulnerable regions in different countries to adopt this model,” Jayaraman claims.
Kalabhavan Jayan, a popular mimicry artist whose home lies along the banks of the Chalakudy River, benefited from early flood warnings issued by the Chalakudy River Protection Forum (CRPF), an organisation specialising in hyper-local advanced weather prediction. In the first week of August this year, following the Wayanad landslide, flooding was reported in many parts of the Chalakudy River basin. “When heavy rains hit our area, I received messages from the CRPF warning of a potential flood. My family and I gathered our belongings and moved to a relief camp. Around 2 a.m., floodwaters from the river entered our home and stayed for about 24 hours,” Jayan recalls, adding that his family was among at least 1,000 households relocated thanks to the timely warnings.
S.P. Ravi, convenor of the CRPF, is working on developing comprehensive flood management and river action plans. He notes that the administration’s responses to their warnings have been mixed. “The Forum informed the district administration about the potential floods three days in advance. Unfortunately, they chose to ignore it. Fortunately, there were no casualties,” says P. Rajaneesh, assistant coordinator of the Forum.
Under their guidance, residents of panchayats such as Pariyaram, Meloor, Karukutti, Parakkadavu, Puthenvelikkara, and the Chalakudy municipality have established a people’s flood monitoring system.
Residents take charge
The emerging community-centric model capitalises on residents’ local knowledge and the use of community-based communication channels, such as WhatsApp groups. This approach enhances coordination between panchayats, civil authorities, and at-risk residents. The network ensures that those in danger receive real-time updates, allowing them to evacuate promptly before conditions worsen.
The floods of 2018 prompted citizens of central Kerala to create the ‘Save Meenachil’ WhatsApp group, which now has over a hundred members, including activists, panchayat representatives, scientists, and residents. “We began by recruiting volunteers to measure rainfall levels using gauges set up in various locations, from the Western Ghats, where the Meenachil river originates, to Vembanad lake, where it flows out,” Eby Emmanuel, secretary of the Meenachil River Protection Committee, says.
“Our initial success came in 2021 when we accurately predicted the likelihood of a flood in Natassery, a settlement near Kottayam. We alerted the authorities using the collected data, enabling them to evacuate the residents,” Emmanuel adds. Today, Emmanuel observes that 60% of the climate volunteers in his group are students. Data is collected daily from 170 rain gauges and 13 river gauges.
Guided by the University
Interestingly, all these local initiatives have partnered with the Advanced Centre for Atmospheric Radar Research at CUSAT for mentoring and guidance. “The real-time rain observations from these groups are gathered and shared in a WhatsApp group that includes authorities, experts, and residents of vulnerable areas. Using this data, we generate forecasts with a 70% accuracy rate for the whole state, enabling potential victims to take appropriate action,” says Abhilash of CUSAT.
“The current official weather forecasting faces many challenges, including the unavailability of government data in the public domain. We also face issues such as low frequency of radar image updates, limited Doppler radar coverage, and erratic weather patterns that make accurate predictions difficult,” Ravi of CRPF notes.
The small village of Puthenvelikkara, located in the Chalakudy-Periyar basin, was completely submerged by floods in 2018, largely due to poor dam management practices. Local activist M.P. Shajan points out that the major dams on the Chalakudy river — Peringalkuthu and Sholayar — are situated in Thrissur district, while the flood-prone areas downstream are in Ernakulam district. This geographical divide often results in a lack of coordination between district authorities.
“Even on days when rainfall exceeds 100 mm, there is no exchange of information regarding flood risks. Warnings about potential floods do not originate from Thrissur. We have two irrigation regulators near Puthenvelikkara, and villages will flood if the shutters are not lifted in time. That’s why we initiated community-level efforts to monitor rainfall levels and share this information with residents. In addition to flood forecasting, we also use the data to predict droughts and inform farming practices,” Shajan points out.
A recent study on India’s southwestern coast shows that localised weather forecasts can significantly benefit small-craft anglers fishing near shore, even during rough offshore weather. The research, carried out jointly by CUSAT, the IMD, and the University of Sussex in the UK, highlights that these localised forecasts help fishers safely launch and land their boats, fish in coastal areas, avoid high-wind conditions in open waters, and return quickly when the weather becomes unfavourable. A hyper-local prediction facility has recently been established along the coast south of Thiruvananthapuram.
Overcoming challenges
Despite growing popularity, hyperlocal climate predictions face challenges. Vishnudas suggests amending the National Disaster Management Act (NDMA) of 2005 to officially incorporate these forecasts for use during extreme weather events. “That would enable district administrations and disaster management authorities to access and use relevant data for mitigation efforts. Currently, officials informally seek out these findings and apply them, but legal recognition would streamline the process and enhance its significance,” he adds.
Vishnudas emphasises the need for automated rain gauge stations in the mountainous areas of the Western Ghats. These stations, which automatically record rainfall data, could eliminate the need for daily visits to remote areas. However, they are costly and require funding. “If the government and civil society collaborate to support the purchase of these stations, it would greatly improve the process and yield better results,” he says.
Grassroots organisations advocate for collaboration and networking with national agencies such as the IMD and the Geological Society of India. This partnership holds the promise of significantly enhancing the accuracy and reliability of disaster predictions. It would be mutually beneficial since these agencies often lack connections to ground-level initiatives and decentralised facilities. On the other hand, grassroots movements would gain access to more accurate predictions with proper mentorship.
Banner image: Chooralmala village in Wayanad after the landslide in August. Repeated extreme weather events like landslides and floods are encouraging citizens in Kerala to take proactive actions to reduce casualties. Image by K.A. Shaji.
Credits
