The influence of precipitation patterns on recent peatland fires in Indonesia インドネシアの泥炭火災に及ぼす降雨パターンの影響


    • Yulianti, Nina



The influence of precipitation patterns on recent peatland fires in Indonesia




Yulianti, Nina


Hokkaido University(北海道大学)








In August 2010, Indonesia formally admitted to producing about 2.1 billion tonnes of CO2, which is a very high amount (as a reference the United States produced about 5.2 billion tonnes of CO2), mostly as a result of peatland fires and deforestation. Although fires have become an annual phenomenon in Indonesia, the rate of rapid deforestation has showed a declining trend (FAO 2010). Indonesia has now been requested to reduce not only carbon emissions from peatland fires, but also haze which is cause to air pollution. Therefore, hotspot data captured by MODIS (Moderate Resolution Imaging Spectroradiometer) from 2002 to 2012 was examined in peatland of Indonesia, Kalimantan, and Sumatra using various grid sizes utilizing latitude and longitude angles from 1° × 1° to 0.01° × 0.01°. Precipitation data was mainly analyzed to explain fire activities in several regions in Kalimantan and Sumatra, simply because precipitation was a common weather data for all the major weather stations in Indonesia. The NOAA (National Oceanic and Atmospheric Administration) and JAMSTEC (Japan Agency for Marine-Earth Science and Technology) definition of El Niño events and their SST anomaly values were also used to analyze their relationship with precipitation and fire activities. Preliminary analysis using a 1° (1° × 1°) grid cell was carried out to locate the most fire-prone areas in Indonesia. Analysis results clearly identified that among the approximately 500 cells covering Indonesia, there were 11 fire-prone cells where the annual average hotspot density exceeded 0.081 hotspots/km2 (= 1,000 hotspots). The top 11 fire prone cells were located solely in Kalimantan and Sumatra; one region (5 adjacent cells) in south Kalimantan, another region (5 adjacent cells) in north Sumatra, and one area (1 cell) in south Sumatra. Detailed analyses using 0.5° × 0.5° grid cells and precipitation data were carried out to understand the seasonal and spatial fire occurrence in Kalimantan and Sumatra more clearly. Most data were tallied every 10 days, enabling a detailed understanding of the seasonal and spatial fire occurrence and precipitation trends. Results of detailed analyses clearly show the number of fire-prone cells with an annual average hotspot density exceeding 0.129 hotspots/km2 (= 400 hotspots) to be 12 for Kalimantan and Sumatra. Two severe fire regions (7 adjacent cells) were identified in the MRP (Mega Rice Project) region in Kalimantan, and in the Dumai region in Sumatra, followed by the Sampit area in Kalimantan and the Palembang area in Sumatra (with 2 adjacent cells). Most fire-prone regions in Kalimantan and Sumatra are located on the peatland and its vicinity. Fire occurrence tendencies in fire-prone regions are mostly explained using the two different precipitation patterns of the region: the summer dry season (SD) pattern and the winter and summer dry season (WSD) pattern. The southern part of Kalimantan and Sumatra, which are located in the southern hemisphere and belong to the SD pattern, show severe fire activities over a relatively longer dry season over a few summer months. The northern part of Sumatra, which is located in the northern hemisphere and accords with to the WSD pattern, shows separate periods of fire activities, due to two dry seasons in both winter and summer months. From a comparison of fire activities in several areas on both islands, it is evident that the most severe peat fires occur in the southern part of Central Kalimantan, due to the relatively longer dry season (of more than 3 months under El Niño) compared with other areas. As a result, a novel method for forecasting and preventing future fires in Kalimantan and Sumatra was proposed. This method involves three factors: the observation of warning fire occurrence by MODIS, the assessment of drought conditions using the accumulated precipitation curve from around early June, and the use of El Niño information from NOAA with the additional from JAMSTEC.


Hokkaido University(北海道大学). 博士(工学)