Responses of biogenic volatile organic compound emissions and surface ozone air pollution in the Greater Bay Area to heatwaves

Abstract

Land cover has significant impacts on local meteorology and biogenic volatile organic compounds (BVOCs) emissions, which in turn affects surface ozone air quality. The Guangdong-Hong Kong-Macau Greater Bay Area (GBA) is a hotspot of ozone air pollution, where BVOCs play an important role. The importance of BVOCs in ozone air pollution strengthens during heatwaves, and heatwaves have been increasing in frequency and intensity in the past two decades. Therefore, it is critical to have accurate BVOCs emissions in atmospheric chemistry models for simulating ozone air quality and supporting the design of efficient ozone air pollution control strategies. However, current BVOCs emissions derived from the outdated and coarse (0.25° x 0.3125° spatial resolution) Community Land Model version 4.5 (CLM4) plant functional types (PFT) map based on Moderate Resolution Imaging Spectroradiometer (MODIS) land cover dataset in 2000 is far from sufficient for producing an accurate BVOCs emission inventory in the GBA. What is more, the CLM4 BVOCs emissions do not consider the substantial land cover change in the GBA in the past two decades, thus incapable of simulating ozone responses to land cover change on a decadal timescale. In this study, we employ the 30m Finer Resolution Observation and Monitoring of Global Land Cover (FROM_GLC2017) map for simulating meteorology with the WRF model and deriving an updated BVOCs emission inventory with MEGAN for September 2017, during which time a one-week heatwave event happened in the GBA. The online two-way coupled regional meteorology-chemistry model WRF-GC v2.0 with the detailed Greater Bay Area anthropogenic emission inventory (GBA-EI) reproduces the MDA8 ozone during non-heatwaves (r = 0.71, NMB = 4.21%), but has significant low biases for MDA8 ozone during heatwaves (r = 0.26, NMB = 27.32%). We find that updating the land cover data from U.S. Geological Survey (USGS) and CLM4 to FROM_GLC2017 is promising to correct the biased low ozone during heatwaves. Compared with the original USGS land cover map, broadleaf evergreen tree forest cover, which has a high BVOCs yield and responds swiftly to temperature increase, increases by 37.29%. Concurrently, the cropland cover, which has a relatively low BVOCs yield, decreases by 44.62%. We will go on to investigate how land cover changes affect ozone trends in the GBA in the past two decades with the long-term FROM_GLC2017 land cover dataset.