EARLY warning systems are only effective when they are linked to resources and effective action on the ground because a warning alone cannot prevent a fire, says Global Environment Centre (GEC)’s peatland programme manager Serena Lew Siew Yan.
“Better prediction can certainly help, but it is not the only factor in reducing peatland fires and haze.
“In many cases, we already know which peatland areas are particularly vulnerable to fires, especially areas that have been drained, degraded, or are frequently affected by human activities.
“What early warning system technologies can provide is more timely information on when those areas are becoming particularly high risk, allowing preventive measures to be prioritised,” she points out.
Lew cautions, however, that a warning is only useful if there is the capacity to act on it.
“This means having effective enforcement, adequate resources, proper peatland and water management, and trained teams on the ground.
“For example, when a high fire risk is identified, authorities and land managers need to have the resources to increase patrols, maintain water levels, prepare firefighting equipment, and respond quickly to any signs of fire.
“Community participation is equally important, particularly because many peatland fires are linked to human activities,” she adds.
Communities, says Lew, need to be involved in monitoring, prevention, and reporting, while landowners and private-sector operators in peatland areas also have an important role in maintaining fire prevention measures within and around their areas of operation, and to also provide support to monitor the buffer zones if possible.
Hence, preventive infrastructure and preparedness measures, she says, need to be in place before high-risk conditions occur.
“For example, installation of canal blocks and other rewetting measures such as activation of tube wells can help maintain or increase soil moisture levels in peatlands and reduce fire risk.
“Maintaining access to tube wells and other water sources can also ensure that sufficient water is available for firefighting efforts by the Fire and Rescue Department Malaysia,” she says.
“Therefore, technology should be viewed as a decision-support tool rather than a standalone solution,” says Lew in an e-mail interview.
Early warning systems, she adds, can provide valuable time for communities and authorities to take preventive actions before a fire starts.
With advance warning, authorities and communities can shift from responding to fires to preventing them, says Lew.
By combining information from satellite observations, drone monitoring, weather conditions, rainfall, peatland water levels, and fire risk indicators, these systems can help identify areas where conditions are becoming increasingly vulnerable to fires.
“Machine learning could further improve this by identifying patterns and predicting areas that may be at higher risk,” she notes.
The greatest benefit comes when early warnings are translated into timely preventive action, supported by appropriate infrastructure, adequate resources, effective enforcement, and active participation from stakeholders, she says, adding that these technologies can be very useful when they are connected to practical peatland management on the ground.
Giving the Internet of Things (IoT)-based environmental monitoring at the Raja Musa Forest Reserve, Selangor, as an example, Lew says it has been used to regularly collect information such as peatland water levels, temperature, and rainfall.
“Changes in these conditions can provide an early indication that an area is becoming drier and increasingly vulnerable to fire.
“This information can support managers in deciding when additional monitoring, water management, or other preventive measures may be needed,” she says.
Set up in 1998 to work on environmental issues of global importance, the GEC is a non-profit organisation working regionally and internationally on a number of programmes, including peatland, where many of the fires occur in Malaysia and Indonesia.
At the regional level, GEC has also facilitated the development of the Asean Fire Alert Tool, which brings together information such as hotspot data from the Asean Specialised Meteorological Centre and the Fire Danger Rating System from the Malaysian Meteorological Department, as well as weather conditions, and wind direction.
Lew says while similar early-warning approaches could be expanded across fire-prone peatlands in South-East Asia, the system should be adapted to the local context, risk profile, and capacity of each area.
“There are already examples in the region that demonstrate how technology can be linked to action,” she says, citing, among others, Indonesia’s Sistem Informasi Perlindungan dan Pengelolaan Ekosistem Gambut and Asap Digital in Jambi where CCTV cameras are used to surveil high risk peatland areas, with information fed through a central command at the local police station.
Another approach focuses on prevention rather than detection: in Thailand and Laos digital tools provide farmers with information on fire risk and whether conditions are suitable for agricultural burning.
“These examples demonstrate that early-warning technology can be scaled across the region, but technology itself is only a part of the solution.
“It needs to be supported by clear mechanisms, trained personnel, enforcement, firefighting resources, and strong stakeholder participation so that warnings are translated into timely action,” she says.
Lew contends that while an early warning system is one important piece of the puzzle, the real impact is when useful information is combined with strong implementation, enforcement, responsible land management, and cooperation between government, communities, and the private sector.
On the Asean Fire Alert Tool, Lew says the wind information is particularly useful because it helps authorities and communities understand the potential direction of smoke movement from areas gutted by fires.
“For instance, ASMC’s wind forecasts can help stakeholders anticipate where smoke haze may potentially travel, enabling them to prepare accordingly,” Lew says, referring to the satellite-based monitoring system used by the Asean Specialised Meteorological Centre.
“Such data does not necessarily predict potential fires, but they provide near-real-time information and indicators of changing fire risk that can support monitoring and decision-making,” she explains.
Lew says in the future, integrating these different datasets with machine learning could potentially improve the ability to identify patterns and forecast areas at higher risk.
However, asked if there are obstacles to further develop the technology, Lew warns there are challenges ahead.
The first is the availability, quality and cost of data.
“Developing a reliable predictive system requires consistent and sufficient detailed data, including weather, rainfall, peatland conditions, hotspot information and historical fire records, and regular and consistent satellite imagery.
“Some of these datasets can also be costly or have restrictions on access and sharing,” she says.
The second challenge is having the technical capacity and resources to manage the system because, says Lew, it is not enough to set up the technology once and leave it running.
“A dedicated team is needed to collect, manage, and analyse the data, maintain and monitor the equipment and systems, interpret the results, and continuously improve the predictive models,” she says.
And, once again, she points out that there must be reaction on the ground.
“Even if a system can accurately identify high-risk areas, the warning will have limited value if there are insufficient resources, personnel or clear procedures to respond.
“Effective enforcement is also important, particularly where fires are associated with illegal or uncontrolled burning,” she says.
Finally, data sharing and timely coordination between different agencies and stakeholders can be challenging, Lew points out.
“Fire and haze management involves government agencies, communities, landowners, and the private sector, and these parties need to be able to share information and respond timely in a coordinated way.
“So the challenge is not simply developing a better prediction model. We need to build the whole system around it – from reliable data and technical capacity, to clear response mechanisms, and timely enforcement.”
