Search

This study investigates the uncertainties associated with estimates of the long-range transport SO2 (LRT-SO2) flow rate calculated hourly using Geostationary Environment Monitoring Spectrometer (GEMS) synthetic radiances. These radiances were simulated over the Korean Peninsula and the surrounding regions using inputs from the GEOS-Chem model for January, April, July, and October 2016. The LRT-SO2 calculation method, which requires SO2 vertical column densities, wind data, and planetary boundary layer information, was used to quantify the effects of the SO2 slant column density (SCD) retrieval error and uncertainties in wind data on the accuracy of the LRT-SO2 estimates. The effects were estimated for simulations of three anthropogenic and three volcanic SO2 transport events. When there were no errors in the GEMS SO2 SCD and wind data, the average true LRT-SO2 flow rates (standard deviation) and those calculated for these events were 1.17 (± 0.44) and 1.21 (± 0.44) Mg/h, respectively. However, the averages of the true LRT-SO2 flow rates and those calculated for the three anthropogenic (volcanic) SO2 events were 0.61 (1.17) and 0.64 (1.20) Mg/h, respectively, in the presence of GEMS SO2 SCD errors. In the presence of both errors in the GEMS SO2 SCD and wind data, the averages of the true LRT-SO2 flow rates and those calculated for the three anthropogenic (volcanic) SO2 events were 0.61 (1.17) and 0.61 (1.04) Mg/h, respectively. This corresponds to differences of 2.1% to 23.1% between the simulated and true mean LRT-SO2 flow rates. The mean correlation coefficient (R), intercept, and slope between the true and simulated LRT-SO2 flow rates were 0.51, 0.43, and 0.45 for the six simulated events, respectively. This study demonstrates that SO2 SCD accuracy is an important factor in improving estimates of LRT-SO2 flow rates. [ABSTRACT FROM AUTHOR]

An aerosol effective height (AEH) retrieval algorithm has been developed using the O4 air mass factor (AMF) at 477 nm from the hyperspectral Ozone Monitoring Instrument (OMI). The magnitude of change in O4 vertical column density (VCD) was topographically and seasonally investigated in Northeast Asia. Its effect on AEH retrieval accuracy has been evaluated using our AEH retrieval algorithm. In addition, the effect of a temperature-dependent cross-section for O4 (TDCS) on AEH retrieval accuracy was quantified. TDCS is found to enhance AEH retrieval accuracy compared with an O4 absorption cross-section at a single temperature of 293 K (SCS), when spatial and seasonal dependency of O4 VCD is applied to the LUT in our algorithm. In comparison between the retrieved AEH and those from ground-based lidar network for the period from January 2005 to June 2009, when both the TDCS and seasonal and topographical O4 VCDs are applied, a Root Mean Square Error (RMSE) is 0.44 km for both smoke and dust types. However, when both a TDCS (SCS) and a single O4 VCD value were applied to the LUT, the RMSE for both aerosol types is 0.52 km (0.51 km). It implies that TDCS contributes most to AEH retrieval accuracy when accurate O4 VCDs are applied to the LUT. For smoke aerosols only, both TDCS and multiple O4 VCD (SCS and single O4 VCD) applications had RMSE value of 0.46 km (0.66 km). Based on synthetic radiances, we also estimated the effects of uncertainties in our algorithm input data such as O4 VCD, TDCS, aerosol type, AOD, and surface reflectance on AEH retrieval error. It was found that large errors are attributable to uncertainties in O4 VCD, AOD, and surface reflectance compared with those of TDCS and aerosol type. Especially, an O4 VCD uncertainty of about 140% caused AEH errors more than 4.3 km. The large improvement of the retrieval performance is expected when the realistic O4 VCD (versus several O4 VCD values in the LUT) is used in the algorithm. [ABSTRACT FROM AUTHOR]