Radiative effect of thin cirrus clouds in the extratropical lowermost stratosphere and tropopause region

Cirrus clouds play an important role in the radiation budget of the Earth; nonetheless, the radiative effect of ultra thin cirrus clouds in the tropopause region and in the lowermost stratosphere remains poorly constrained. These clouds have a small vertical extent and optical depth, and are frequently neither observed even by sensitive sensors nor considered in climate model simulations. In addition, their shortwave (cooling) and longwave (warming) radiative effects are often in approximate balance, and their net effect strongly depends on the shape and size of the cirrus particles. However, the CRyogenic Infrared Spectrometers and Telescopes for the Atmosphere instrument (CRISTA-2) allows ultra thin cirrus clouds to be detected. Here we use CRISTA-2 observations in summer 1997 in the northern hemisphere midlatitudes together with the Suite Of Community RAdiative Transfer codes based on Edwards and Slingo (SOCRATES) radiative transfer model to calculate the radiative effect of observed ultra thin cirrus. Using sensitivity simulations with different ice effective particle size and shape, we provide an estimate for the uncertainty of the radiative effect of ultra thin cirrus in the extratropical lowermost stratosphere and tropopause region during summer and - by extrapolation of the summer results - for winter. Cloud top height and ice water content are based on CRISTA-2 measurements, while the cloud vertical thickness was predefined to be 0.5 or 2 km. Our results indicate that if the ice crystals of these thin cirrus clouds are assumed to be spherical, their net cloud radiative effect is generally positive (warming). In contrast, assuming aggregates or a hexagonal shape, their net radiative effect is generally negative (cooling) during summer months and very likely positive (warming) during winter. The radiative effect is in the order of +/-(0.1-0.01) W/m2 for a realistic global cloud coverage of 10%, similar to the magnitude of the contrail cirrus radiative forcing (of ~0.1 W/m2). The radiative effect is also dependent on the cloud vertical extent and consequently the optically thickness and effective radius of the particle size distribution (e.g. effective radius increase from 5 to 30~microns results in a factor ~6 smaller long and shortwave effect respectively). The properties of ultrathin cirrus clouds in the lowermost stratosphere and tropopause region need to be better observed and ultra thin cirrus clouds need to be evaluated in climate model simulations. --