Measurements are presented of the wavelength dependence of the aerosol absorption coefficient taken during the Tropical Aerosol Radiative Forcing Observational Experiment (TARFOX) over the northern Atlantic. The data show an approximate Ϫ1 variation between 0.40 and 1.0 m. The theoretical basis of the wavelength variation of the absorption of solar radiation by elemental carbon [or black carbon (BC)] is explored. For a wavelength independent refractive index the small particle absorption limit simplifies to a Ϫ1 variation in relatively good agreement with the data. This result implies that the refractive indices of BC were relatively constant in this wavelength region, in agreement with much of the data on refractive indices of BC. However, the result does not indicate the magnitude of the refractive indices. The implications of the wavelength dependence of BC absorption for the spectral behavior of the aerosol single scattering albedo are discussed. It is shown that the single scattering albedo for a mixture of BC and nonabsorbing material decreases with wavelength in the solar spectrum (i.e., the percentage amount of absorption increases). This decease in the single scattering albedo with wavelength for black carbon mixtures is different from the increase in single scattering albedo for most mineral aerosols (dusts). This indicates that, if generally true, the spectral variation of the single scattering albedo can be used to distinguish aerosol types. It also highlights the importance of measurements of the spectral variation of the aerosol absorption coefficient and single scattering albedo.

No abstract

SUMMARYCalculated irradiances from a new radiation code are compared with in situ observations of short-wave irradiances from the UK Meteorological Office's C-130 aircraft. Three cases of clear skies are studied and four where a liquid-water boundary-layer cloud was present. Under clear-sky conditions the modelled and in situ observations agree to within 3%, which is the estimated accuracy of the observations. In the cloudy-sky cases the albedo and transmittance agree to within fO.l but the absorption in the model is higher than that observed, sometimes by a factor of two; there is no evidence of anomalous absorption in the observations. The observed absorptions do not exceed 6% for the stratocumulus cases considered. The results clearly identify the problems of representing inhomogeneous clouds as plane parallel layers in radiation models. Analysis of the variability of the cloud microphysics provides some insight into the importance of regions of low optical depth within the clouds.

We present details of a scheme for retrieving cirrus cloud optical thickness and effective particle size from nadir‐viewing reflectance measurements. Two near‐infrared wavelengths are used, one with negligible ice absorption (1.04 μm) and one with significant ice absorption (1.55 μm). Four ice crystal shapes are used in the scheme: ice spheres, hexagonal columns, hexagonal plates and randomized polycrystals. We highlight the sensitivity of the retrievals to the shape of the phase function, which follows from the particular ice crystal shape assumed. Five aircraft‐based retrievals are presented, all cases occurring during the 1993 intensive field campaign of the European Cloud Radiation Experiment (EUCREX'93). The retrieved effective sizes are compared with the in situ measurements taken from the same cases. In these comparisons, account has to be taken of the fact that (a) in situ replicator data indicate that the presence of small (<100 μm) ice crystals can have a significant effect on the effective crystal size, and (b) the measured crystal sizes consistently show a systematic decrease with height in all cases, the effective sizes at cloud base being typically two or three times larger than at cloud top. Taking these considerations into account, we conclude that the assumption of polycrystals gives the most consistent agreement with the in situ measurements.

No abstract