Characterizing the radiative effects of smoke from large scale vegetation fire events using radiometric surface observations, satellite retrievals and trajectory modeling

In this paper, we present the results of a collaborative study to assess the direct radiative effect of smoke from large biomass burning events on the surface insolation. We report the preliminary results of a 1986 fire season study which is centered in the Republic of Congo in Africa and the results of a study focusing on a large boreal forest fire event during July 1989 in Manitoba, Canada.

Konzelman et al. (1996) observed large differences between predicted all-sky solar irradiance from the World Climate Research Program (WCRP) solar Surface Radiation Budget (SRB) data set (Whitlock et al., 1995) and surface insolation measurements at 4 sites in Zaire (now the Republic of Congo). The differences corresponded with the peak in the burning season during that year. We analyze the surface radiometer daily averaged time series during this time from several surface radiometer sites in the vicinity of the burning by calculating the difference between the measured flux in the clear-sky and a "clean-sky" (no aerosol, no cloud) radiative flux. This solar residual flux (SRF) will be computed for the time series before and after the burning episodes. The SRF is sensitive to changes in the background aerosol loading and these fire episodes easily provide enough loading to exceed the uncertainties associated with the measurements. A similar analysis is presented for surface radiometer sites downwind of a large boreal forest fire outbreak in 1989. The SRF fluxes are computed during this case as well. Early results have indicated that during the period of the fires clear-sky radiative fluxes over Madison, Wisconsin and the island of St. Pierre and Miquelon (located off the coast of Newfoundland, Canada) showed differences between estimated and measured daily averaged clear-sky solar insolation of more than 80 W m^-2.

Owing to the large uncertainties associated with solar irradiance measurements, the NASA LaRC Trajectory Model (LTM) initialized with atmospheric profile information from the European Community Medium-Range Weather Forecasting Reanalysis (ERA-15) to trace origins of aerosol parcels over the set of selected sites. The locations of fires are given by previous DMSP analysis and by new AVHRR analysis (Baum and Trepke, 1999; Cox et. al., this conference). Since the LTM is fully 3-D these probabilities are also analyzed in terms of a profile using simple assumptions concerning boundary layer mixing and scavenging. To complement this data set, we compare the surface insolation measurements during this time period to newly derived solar insolation estimates from the WCRP Global Energy and Water Cycle Experiment (GEWEX) SRB data set at 1 degree resolution. The LTM is used to derive the probability that parcels containing smoke are being advected over the site to collaborate the radiometric analysis.