| description abstract | The 2D/3D cloud model complex with explicit microphysics and radiation described in Part I is used to simulate the development of a midlatitude cirrus cloud, including interaction with radiation. To account for the effects of the interaction of various scales of motion on cloud development, a synoptic-scale vertical velocity field is superimposed on the mesoscale velocity field generated by the model, mimicking the effects of an upper-level shortwave trough. The main results under the conditions simulated here are the following. Cirrus cloud growth is much slower than assumed previously, because the process of vapor deposition to ice crystals is far from instantaneous: the crystal phase relaxation time (i.e., the characteristic time of vapor absorption by crystals) takes 0.5?2.0 h. Even after 1 h of cloud development, supersaturation with respect to ice can remain 5%?10%, while the condensed ice is only 40%?60% of the amount that would be realized assuming that all excess vapor is transformed into ice in typical model time steps. Although experimental and theoretical studies have produced widely divergent longwave mass absorption coefficients αabsm, ranging from 100 to 3500 cm2 g?1, model results show that a single ?representative? value of αabsm is inappropriate. Vertical profiles typically exhibit values of ?800?1000 cm2 g?1 in the upper cloud region containing the smallest particles, in contrast to ?100?300 cm2 g?1 for the larger crystals in the main cloud. The optical scattering coefficients behave similarly, with typical values of ?2000?2500 cm2 g?1 in the upper cloud regions and ?300?500 cm2 g?1 in the lower cloud regions. A strong horizontal variability is also a characteristic feature of these coefficients. Many GCM and climate models use seemingly overestimated αabsm values (e.g., 1000 cm2 g?1). Sensitivity tests show that the use of such values increases cooling in the upper cloud and heating in the lower cloud, which can lead to an unwarranted increase in upper-tropospheric static instability. The postulated effects of the positive feedbacks between clouds and greenhouse gas?induced global warming would likely be different in magnitude (or in sign) if the more realistic approach of using cloud microstructure?dependent absorption and scattering coefficients could be adopted. Consideration of microphysics also shows that the decrease in the shortwave radiative balance (albedo effect) in the simulated midlatitude cirrus cloud exceeds the net gain in the longwave balance (greenhouse effect) near midday, due to the abundance of relatively small crystals in the upper cloud region where cloud regeneration is taking place. | |
