Abstract. Fires affect the composition of the atmosphere and Earth's
radiation balance by emitting a suite of reactive gases and particles. An
interactive fire module in an Earth system model (ESM) allows us to study
the natural and anthropogenic drivers, feedbacks, and interactions of open
fires. To do so, we have developed pyrE, the NASA GISS (Goddard Institute for Space Studies) interactive fire
emissions module. The pyrE module is driven by environmental variables like
flammability and cloud-to-ground lightning, calculated by the GISS ModelE
ESM, and parameterized by anthropogenic impacts based on population density
data. Fire emissions are generated from the flaming phase in pyrE (active
fires). Using pyrE, we examine fire occurrence, regional fire suppression,
burned area, fire emissions, and how it all affects atmospheric composition.
To do so, we evaluate pyrE by comparing it to satellite-based datasets of
fire count, burned area, fire emissions, and aerosol optical depth (AOD). We
demonstrate pyrE's ability to simulate the daily and seasonal cycles of open
fires and resulting emissions. Our results indicate that interactive fire
emissions are biased low by 32 %–42 %, depending on emitted species,
compared to the GFED4s (Global Fire Emissions Database) inventory. The bias in emissions drives
underestimation in column densities, which is diluted by natural and
anthropogenic emissions sources and production and loss mechanisms.
Regionally, the resulting AOD of a simulation with interactive fire
emissions is underestimated mostly over Indonesia compared to a simulation
with GFED4s emissions and to MODIS AOD. In other parts of the world pyrE's
performance in terms of AOD is marginal to a simulation with prescribed fire
emissions.