Crowdsourced hotspot validation and data visualisation for location-based haze mitigation

ABSTRACT

Haze over Sumatera and Kalimantan has been a prolonged trans-boundary issue in South East Asia mainly due to setting fire to drained peatland. At present, fire sources (i.e. hotspots) are located based on satellite data. Sensors such as MODIS and AVHRR detect extremes in average temperatures of an area. The hotspots have low spatial resolution and large spatial footprints, thus making it harder to detect fires. This research proposed a ground-based spatial validation of satellite data based on crowdsourcing in order to obtain more accurate estimates of the location and severity of the fire. We developed an Android application for reporting and validating fires in peatlands. Crowd data collected were integrated with satellite hotspot data by the dashboard system as a monitoring platform for government agencies. The 110,888 hectares of Padang Island, in Riau Province, were chosen as the study area given its vulnerability to peatland fire and imminent danger of subsidence as the collateral effect of draining peatlands. Residents of Padang Island tested the use-case scenario of the app to assess its applicability. The study showed the potential use of mobile apps for local communities to help the government validate hotspots for haze mitigation.