资源环境科技发展态势分析平台

We examine upcoming geostationary satellite observations of formaldehyde (HCHO) vertical column densities (VCDs) in East Asia and the retrieval sensitivity to the temporal variation of air mass factors (AMFs) considering the presence of aerosols. Observation system simulation experiments (OSSE) were conducted using a combination of a global 3-D chemical transport model (GEOS-Chem), a radiative transfer model (VLIDORT), and a HCHO retrieval algorithm developed for the Geostationary Environment Monitoring Spectrometer (GEMS), which will be launched in 2019. Application of the retrieval algorithm to simulated hourly radiances yields the retrieved HCHO VCDs, which are then compared with the GEOS-Chem HCHO VCDs as true values for the evaluation of the retrieval algorithm. In order to examine the retrieval sensitivity to the temporal variation of AMF, we examine three AMF specifications, AMF(m), AMF(h), and AMF(mh), using monthly, hourly, and monthly mean hourly input data for their calculation, respectively. We compare the retrieved HCHO VCDs using those three AMFs and find that the HCHO VCDs with AMF(h) are in better agreement with the true values than the results using AMF(mh) and AMF(m). AMF(mh) reflects diurnal variation of planetary boundary layer and other meteorological parameters, so that the results with AMF(mh) show a better performance than those with AMF(m). The differences between AMF(h) and AMF(m) range from -0.76 to 0.74 in absolute value and are mainly caused by temporal changes in aerosol chemical compositions and aerosol vertical distributions, which result in 27 to 58 and -34 to 43% changes in HCHO VCDs over China, respectively, compared to HCHO VCDs using AMF(m). We apply our calculated AMF table together with OMI aerosol optical properties to OMI HCHO products in March 2006, when Asian dust storms occurred, and find -32 to 47% changes in the retrieved HCHO columns due to temporal changes in aerosol optical properties in East Asia. The impact of aerosol temporal variability cannot be neglected for future geostationary observations.