Modeling study of PM2.5 pollution episode of early spring 2019 in Hokkaido, Japan caused by biomass burning in Northeast China

Abstract

Biomass burning (BB), in particular agricultural waste burning (Agri-BB), occurs at random locations, scales, and times. These factors make it challenging to detect Agri-BB accurately through satellite observations. Thus, the BB emission inventories using satellite observation data have uncertainties for their emission estimation approach and cause poor model performance for air pollutants including PM2.5. We utilized the two BB emission inventories, GFEDv4.1s and FINNv2.5 with the CMAQ model to simulate the PM2.5 heavy pollution episode in Hokkaido 2019. To estimate Agri-BB contributions, we conducted three simulation cases for each BB emission inventory: with and without Agri-BB emission, and the boosted Agri-BB emission cases. The baseline simulation failed to capture the temporal and spatial variation patterns of PM2.5. Meanwhile, the boosted Agri-BB case could show favorable performance for PM2.5 concentrations. These results indicated that the two BB emission inventories underestimated Agri-BB emissions. In the two boosted Agri-BB cases, the PM2.5 contributions from Agri-BB accounted for more than 50% during the episode. Moreover, high PM2.5 emissions were found in Northeast China and its surrounding regions similar to the two boosted Agri-BB cases. Consequently, the results revealed that Agri-BB emissions during the episode were significantly derived from the agricultural areas in Northeast China.