Northern peatlands under fire: Projecting smouldering combustion loss in an uncertain future with implications for atmospheric metal emissions

Peatland ecosystems, while covering only ~3% of the land surface area, are globally-important sinks of atmospheric carbon dioxide and regionally-important sinks of pollutants such as toxic metals and metalloids. While metal concentrations in peatlands are generally low, concentrations can be far higher near current and historic industrial centres, particularly in the upper few decimetres of the peat profile. Under normal conditions these metals remain safely sequestered in the peat. However, there is concern that, in addition to direct carbon emissions, peatland wildfires could represent a major pathway for metal mobilization and transport. Moreover, peat fires are dominated by smouldering, which is a low temperature combustion that leads to high concentrations of particulate matter within the smoke, representing a major health risk for communities impacted by wildfire smoke plumes.

 

With projected future climate change, annual area burned and subsequent carbon emissions are expected to rise, with drastic increases associated with high climate-forcing scenarios. In addition to greater area burned, higher evaporative losses associated with warming conditions may lead to increased peat smouldering vulnerability during wildfire. However, differences in local climate, projected changes in precipitation, and peatland type may have strong regionally-dependent mitigating effects.

 

Using the MODIS burned area product, we first develop an empirical relationship between current average area burned and regional climate. Using the climate-driven relationship, we estimate future changes in area burned from multiple climate models (CMIP6 GCMs) and across several climate-forcing scenarios (SSP 2-4.5, 3-7.0,and 5-8.5). In addition to changes in area burned, peat smouldering carbon loss is evaluated by simulating peat moisture profiles using HYDRUS-1D within a phase-space defined by evaporative demand and water table (WT) position. Under contemporary conditions, peat smouldering loss is concordant with the depth of peat that exceeds a critical soil water tension threshold under steady state conditions using the mean WT position. The impacts of climate change on smouldering carbon loss is then estimated based on the change in position within the evaporative demand–WT phase space. Peatland WT sensitivity to temperature and precipitation are taken from the literature and used to estimate changes in WT based on GCM projections across SSPs. Combined with published data on peatland type and location, we produce some of the first ever hemisphere-wide estimates of northern peatland carbon loss from smouldering due to climate change. 

 

These spatially explicit results are used to highlight regions of overlap between increased burn area and severity with areas of high peat metal contamination. Taken together with estimates of particulate emissions from smouldering combustion, we provide an estimate of how climate change may increase global particulate matter emissions from northern peatlands. Moreover, uncertainty in the empirical relation and inter-model variability are used to quantify the confidence intervals for both projected area burned, peat burn severity, and thus particulate matter emissions, in order to highlight where future efforts are best focused for improving robustness of future projections.