Modeling Burned Areas in Indonesia: The FLAM Approach

Abstract

Large-scale wildfires affect millions of hectares of land in Indonesia annually and produce severe smoke haze pollution and carbon emissions, with negative impacts on climate change, health, the economy and biodiversity. In this study, we apply a mechanistic fire model to estimate burned area in Indonesia for the first time. We use the Wildfire Climate Impacts and Adaptation Model (FLAM) that operates with a daily time step on the grid cell of 0.25 arc degrees, the same spatio-temporal resolution as in the Global Fire Emissions Database v4 (GFED). GFED data accumulated from 2000–2009 are used for calibrating spatially-explicit suppression efficiency in FLAM. Very low suppression levels are found in peatland of Kalimantan and Sumatra, where individual fires can burn for very long periods of time despite extensive rains and fire-fighting attempts. For 2010–2016, we validate FLAM estimated burned area temporally and spatially using annual GFED observations. From the validation for burned areas aggregated over Indonesia, we obtain Pearson’s correlation coefficient separately for wildfires and peat fires, which equals 0.988 in both cases. Spatial correlation analysis shows that in areas where around 70% is burned, the correlation coefficients are above 0.6, and in those where 30% is burned, above 0.9.

Sea level rise and climate change acting as interactive stressors on development and dynamics of tropical peatlands in coastal Sumatra and South Borneo since the Last Glacial Maximum

“…Despite the potential natural origin of fires (Hapsari et al, 2021; Krasovskii et al, 2018), the presence or increase in charcoal or fire are often inferred to result from early anthropogenic activities (Anshari et al, 2001; Hope et al, 2005; Yulianto et al, 2004). To be able to distinguish between the two possible drivers, other evidence is required like, for example, archaeological or historical records, or an indication of anthropogenic disturbance that is commonly accompanied by pollen of agricultural taxa or a large quantity of pollen of open vegetation taxa (Hapsari et al, 2021).…”

“…In addition to other factors (see Section 4.3), the Mid‐Holocene ENSO possibly played a role in enhancing dry fuel (aboveground biomass) availability during dry El Niño events (Hapsari et al, 2021). Moreover, lightning hazards, a natural ignition source in tropical humid forests (Krasovskii et al, 2018; Tutin et al, 1996), increased during El Niño events over land and coasts in Indonesia (Dowdy, 2016; Hamid et al, 2001). During El Niño, an increase in air surface temperature and a larger sea‐land temperature contrast can trigger the formation of deep convective storms, thus, produce more lightning discharge (Hamid et al, 2001).…”

A Neural Network Model for Wildfire Scale Prediction Using Meteorological Factors

“…The meteorological variables were the maximum temperature (MT), cooling degree days (CDD), heating degree days (HDD), total rain (TR), total snow (TS), snow on ground (SG), direction of maximum wind gust (DMG), and speed of maximum wind gust (SMG). The traditional method for rating wildfire scale is based on the size of the burned area [39]. However, we propose a more meaningful definition of scale that encompasses not only the size of the burned area but also the fire's duration [40].…”

Opportunities to Optimize the Palm Oil Supply Chain in Sumatra, Indonesia

“…A 25 km × 25 km grid size was chosen to match the resolution of a previous work on forest fires in Indonesia [31], the results of which we intend to merge with the results of this work to serve as input for a harmonized risk assessment study in the future. The grid size gives a total amount of 656 grid points in Sumatra Island, which is a relevant number for the model to be resolved in a reasonable amount of time without losing accuracy in the results.…”

“…The economic losses from peat fires were based on 41.73 USD/ha/year [52]. We estimate that 39,000 ha are affected [31]. The water supply disruptions were calculated by multiplying the total number of oil palm plantations with the amount of water lost for every one ha of oil palm plantation developed (500 m 3 /ha/year) and water price (0.402 USD/m 3 ) [52].…”