Wildfire risk and emissions predictions for SE Asia peatlands based on mechanistic modeling

Abstract:

South East Asian peatlands, which can reach 20m in depth, store approximately 100 billion tons of carbon. The vast majority of this carbon is located below the water table and is largely protected from decomposition and release to the atmosphere. However, natural and anthropogenic droughts lower the water table, expose carbon to decomposition, and increase the risk of wildfire which burns the peat. The 1997/8 El Niño caused a severe drought that led to peatfire emissions of 1.8 billion tons of carbon, roughly 25% of the total anthropogenic emission for the year. Anthropogenic droughts, via the draining of peatlands for oil palm, acacia, and rice plantations, pose enormous risk for current and future carbon emissions. In this study, we use a new peatlands version of the Ecosystem Demography (ED2) model to predict the soil organic carbon that is susceptible to decomposition and fire under drought conditions, both natural and those due to man-made drainage ditches. Driven with historic weather data, ED2-peatlands is able to reproduce interannual and seasonal variation in the water table. We evaluate the model’s predictions of soil moisture against estimates of remotely-sensed surface soil moisture from the Soil Moisture Active Passive (SMAP) satellite. This model has vertically resolved soil carbon and can quantify the carbon content of peat above the water table that is vulnerable to fires and decomposition. The simulated vulnerable carbon correlates with MODIS fire hotspot data over the past decade indicating that fires are more likely in areas of dry peat. Consequently, large amounts of tropical peatland carbon are vulnerable to being lost if they are drained. Simulations of drained peatlands will also be presented quantifying the additional risk of carbon emissions from fire and decomposition associated with converting natural peatlands to agriculture.