Investigating air quality and climate impacts of fires using an interactive fire module in CESM2

Fires and associated emissions of trace gases and aerosols are an important component of Earth System Models that influence air quality and climate. These emissions are strongly dependent on various processes, including climate variability and human impacts. However, fire emissions in Earth System Model simulations are often prescribed, limiting the analysis of the driving processes and their interactions. Here, we use the Community Earth System Model version 2 (CESM2), and run a fully coupled historical simulation (1850 to the present) with the interactive fire module in the Community Land Model (CLM). We compare this to a second simulation using prescribed fire emissions from the Coupled Model Intercomparison Project phase 6 (CMIP6). The initial results show that even though the simulation with the interactive fire module performs well in the present day, there is a high bias in the burned area from 1900-1950, compared with the simulation using prescribed fire emissions. The bias in the burned area results in larger biomass burning emissions. For example, the simulation with the interactive fire module has much higher fire black carbon emissions, resulting in lower surface temperature globally and a reduced temperature trend.

To understand the differences between interactive and prescribed fire simulations, we analyze trends in the drivers of fires (e.g., lightning and soil wetness), fire behaviors and trends in different fire types in the CLM fire module (namely agricultural fires, deforestation fires in the tropics, peat fires, and others). We also evaluate the components and assumptions in the CLM fire module such as fire occurrence, burned area, and fire duration in the present-day condition. We compare emissions of trace gases and aerosols from the CLM fire module with the Fire INventory from NCAR (FINN), an observation-based fire emission inventory. This analysis will help identify drivers for these inconsistencies. Finally, after understanding the inconsistencies of the simulations, and evaluating the components and assumptions in the module, we study air quality and climate impacts of fires through our fully coupled historical simulations. Besides understanding historical fire impacts, this study will also be helpful to understand fire behaviors and changes in the future projection.