Atmospheric conditions are well represented by the laboratory experiments. This is proven by various parameters identified (e.g., emission factors, single scattering albedo).

Experiments were carried out under reproducible conditions for a variety of different fuel types (African vegetation: acacia, musasa, mupangara, Mediterranean vegetation: aleppo pine, kermes oak, boreal vegetation: oak, spruce, pine; partly with dry or green material and Indonesian as well as German peat).

Peat results deviate from other fuel types (e.g., hygroscopicity, morphology of the particles, DCO/DCO2-ratio, size-resolved chemistry).

Fig.1: Single particle images from Indonesian peat (left) and oak burnt (right) taken with electron microscopical techniques (subproject P5).
Fig. 2: Hygroscopic growth for selected aerosol particle sizes for Northern German peat and oak (subproject P1).
Fig. 3: CCN size distributions for various supersaturations (left: oak, right: Northern German peat): significant reduction of large CCN in peat smoke (subproject P4).

Fig. 4 a and b: Mass fractions of chemical species in smoke from Indonesian peat (A) and Northern German peat (B) from Berner impactor sample. (Stage 1: 0.05-0.14 µm, 2: 0.14-0.42 µm, 3: 0.42-1.2 µm, 4: 1.2-3.5 µm, 5: 3.5-10µm) (subproject P1).

Fig. 5: Aerosol particle size distributions for different biomass fuel (subproject P1).

Warm rain is suppressed due to the excess supply with aerosol particles. Precipitation formation via the ice phase is possible because of the large insoluble fraction of the particles.

Direct comparability of radiation modelling and experimental results allows for a 'closure': The temporal development of optical properties cannot be attributed to the changed particle size distributions only, change of the chemical composition is required.

Convective modelling of the Chisholm Fire (Canada): Meteorology (especially wind direction) highly affects the plume shape and height.

REMO: Reasonable emission estimates led to realistic results over Indonesia with respect to atmospheric aerosol burden and precipitation changes.

The Convective Cloud Field Model (CCFM) allowing a sub-gridscale cloud evolution was developed for use in global climate models.

Fig. 6: Liquid and ice water contents in g/kg as functions of temperature for a warm case and for different insoluble particles in immersion freezing (left side) and contact freezing (right side) (subproject P2).

Fig. 7: Comparison between measured
and modelled single scattering albedo
(SSA) for a 1-hour fire with dry oak as
fuel (subprojects P3 and P4).

Fig. 8: Influence of the mixing type assumption of organic (OC) and black carbon (BC) on the single scattering albedo at 550 nm (average size distributions from measurements of an oak fire) (subproject P3).
Fig. 9: Evolution of a convective plume depending on wind direction simulated with ATHAM (subproject P6).
Fig. 10: Vertical structure of the convective heating rates. In contrast to the standard description (left), use of the CCFM (right) leads to a vertical structure of the convective heating rates in a GCM, representing three different cloud types (subproject P8).

Fig. 11: REMO model results: convective precipitation [mm/6h] over Indonesia, March 1 1998, 6 UTC: suppression of convective precipitation by smoke in the interaction run (subproject P7).

Fig. 12: REMO model results: Delta precipitation, Sept. 1997, Indonesia, 'interaction' minus 'control' run (%) per interval [0.5 mm/6h]: redistribution of rainfall
(subproject P7).
Last change: 2004/06/25