Moving Toward Continuous Satellite Monitoring of PM2.5 Using the GOES Aerosol/Smoke Product (GASP) and Aircraft Profiles

The NOAA Geostationary Operational Environmental Satellite (GOES) makes measurements of aerosol optical depth (AOD) every 30 minutes during daylight hours. Those measurements then feed the Automated Smoke Detection and Tracking algorithm that uses fire counts, trajectory modeling and pattern recognition to identify fire plumes, especially in the western U.S. Tying these satellite measurements to surface measurements of fine particles (PM2.5) would be a considerable benefit to the air quality community and to people who live in areas with elevated fine particle concentrations. Currently, these retrievals are useful in identifying areas of elevated PM2.5 concentrations and in forecasting PM2.5 by federal, state and local agencies, but are largely limited to qualitative measures of fine particle loading. Among other issues, layers of fine particles well above ground level, cloud contamination, and particle growth by addition of water in areas of high relative humidity are examples of barriers to a direct relationship between surface PM2.5 and satellite AOD. We have identified a path forward by using aircraft profiles to determine the vertical distribution of aerosol scattering in the atmosphere. In addition, long-term measurements of scattering and rapid measurements of PM2.5 at ground-based field sites have provided a relationship between scattering and mass. Simultaneous measurements of relative humidity and temperature allow one to calculate scattering the particles would have in a dry environment and relate that to fine particle mass measurements. A relationship between rapid scattering measurements and much slower PM2.5 mass measurements is then developed, which is used to tie rapid aircraft measurements of scattering to mass. In turn, aircraft profiles are then used to tie column measurements to those at the ground and to identify cases when satellite retrievals are likely to fail. The resulting algorithm should apply throughout much of the eastern U.S., so long as the aerosol types are similar to the fine, secondary industrial aerosol common to the area where the relationship was developed. The method used to develop this algorithm should prove to be useful over a larger area, though the measurements themselves would have to be repeated elsewhere. Ultimately, we aim to produce estimates of surface PM2.5 for much of the time when GOES retrievals of AOD are available and a screening process to indicate when a direct relationship cannot be determined.