Figure 4. Latitudinal variations of d ln t/dT (a), d ln re/dT (b), and dTc/dT (c) obtained for low-level, single-layer clouds over the Pacific Ocean between 55° S and 55° N. The results for overcast and threshold cloudy pixels are compared.

The sensitivities to the atmospheric temperature at 740 hPa in Fig. 4 are also compared between the overcast (solid circles) and threshold cloudy pixels (open circles). It is seen from the figure that the negative t-T relationships are mostly found in midlatitudes and appear to be relatively unaffected by the choice of overcast and partly cloudy pixels, even though partly cloudy pixels have significantly smaller t than the overcast pixels. There are positive t-T relationships found in the tropics, and the differences between overcast and threshold cloudy pixels are also found to be larger in the tropics. The positive or negative t-T relationship may be determined by the different properties of clouds in different regimes, which requires more detailed investigations.

To examine the sensitivity of L, two approaches are used here: (1) the moist adia-batic liquid water content as documented by the theoretical study of Betts and Harshvandhan (1987) for low clouds (>680hPa), and (2) the subadiabatic liquid water content as documented by the observational study of Somerville and Remer (1984). Figure 5 shows the best-fit regression for the values of d ln L/dTc derived for the moist adiabatic (open points; data taken from Betts and Harshvandhan, 1987) and for the subadiabatic (solid points; data taken from Somerville and Remer, 1984). The regression for the moist adiabatic is given by d ln L dTr.

and for the subadiabatic it is given by

0 0

Post a comment