Uncertainties of wild-land fires emission in AQMEII phase 2 case study: Fid-Bau Portal

Uncertainties of wild-land fires emission in AQMEII phase 2 case study

Soares, J. / Sofiev, M. / Hakkarainen, J.

Abstract The paper discusses the main uncertainties of wild-land fire emission estimates used in the AQMEII-II case study. The wild-land fire emission of particulate matter for the summer fire season of 2010 in Eurasia was generated by the Integrated System for wild-land Fires (IS4FIRES). The emission calculation procedure included two steps: bottom-up emission compilation from radiative energy of individual fires observed by MODIS instrument on-board of Terra and Aqua satellites; and top-down calibration of emission factors based on the comparison between observations and modelled results. The approach inherits various uncertainties originating from imperfect information on fires, inaccuracies of the inverse problem solution, and simplifications in the fire description. These are analysed in regard to the Eurasian fires in 2010. It is concluded that the total emission is likely to be over-estimated by up to 50% with individual-fire emission accuracy likely to vary in a wide range. The first results of the new IS4FIRESv2 products and fire-resolving modelling are discussed in application to the 2010 events. It is shown that the new emission estimates have similar patterns but are lower than the IS4FIRESv1 values.

Highlights • Main uncertainties of wild-land fire emission estimates is discussed. • Total emission can be over-estimated up to 50% with individual-fire emission accuracy. • IS4FIRESv1 emissions in Europe are over-estimated in-average by 20–30%. • Impact on total emissions probably comes from under-stated injection height. • High-energy sources mis-interpreted by MODIS as fires bring about a few tens of %.