14.B5 Upgrading the Cloud Optical Properties in the UCLA-AGCM

1. Introduction

Clouds strongly influence tropospheric radiative heating rates and surface radiative fluxes. Yet there are considerable uncertainties in determining the cloud optical properties in the framework of atmospheric general circulation models (AGCMs). The optical thickness of high-level clouds varies both spatially and temporally in a range of at least from 0.1 to 20 (see Liao et al. 1995 and Rossow and Schiffer 1991 for SAGE II and ISCCP C2 data respectively).

Ma et al. (1994) and Yu and Mechoso (1997) have found strong sensitivities of the simulation of sea surface temperatures with a coupled ocean-atmosphere GCM to changes in the emissivity of high clouds. This paper examines the sensitivity of an AGCM to successively greater sophistication in the specification of long-wave (LW) and short-wave (SW) cloud optical properties with focus on the latter. In particular, we demonstrate the role radiation can play in the formation of stratus clouds in the framework of a well-mixed planetary boundary layer (PBL). To quantify the impact of stratus and high clouds on the radiative fluxes in the AGCM we perform experiments with a one-column version of the AGCM.

2. Cloud Optical Properties in AGCMs

The following three parameterizations of cloud optical properties used in AGCMs are compared in this study: Katayama (1972) specifies the SW cloud optical properties as a function of cloud layer depth and height ("one cloud type" or "1CT"). Harshvardhan et al. (1989) use a function of cloud layer depth, cloud temperature and cloud type (optically thin large-scale supersaturation "stratiform" clouds and optically thick "cumuliform" anvils -- "two cloud types" or "2CT"). Köhler et al. (1997) use predicted cloud water to calculate the cloud optical properties ("continuous cloud type" or "CCT").

3. One-column Model Experiments

Tests using a one-column version of the UCLA AGCM acting on a typical warm pool profile with high clouds showed that the net surface flux below high clouds can vary up to a factor of two between the two cloud types when using the "2CT" cloud optical properties. Stratiform high clouds have optical thicknesses around 0.1 and affect the surface SW flux and the atmospheric heating very little. The assigned SW transmissivity in "1CT" lies just in the middle of the two "2CT" extremes, but the absorptivity is below both "2CT" values (Figure 1a).

Experiments with the one-column model acting on a profile typical of stratus-cloud covered regions have shown a two times larger absorption of SW radiation in stratus clouds when using "2CT" cloud optical properties. Also, the net surface SW flux was about 35% larger when using "2CT". Both effects contribute to a heating of the planetary boundary layer (PBL) over land (Figure 1b).

Figure1. Short-wave heating profiles calculated from a one-column version of the UCLA-GCM using typial warm pool high-cloud (a) and off the coast of California stratus (b) temperature and humidity profiles. Heating profils were separately calculated for the 1CT and 2CT (thin and thick for high clouds) SW parameterizations.The vertical axis represents the model layers.

4. Full AGCM Experiments

4.1. From one to two cloud types

Simulations using the full UCLA AGCM with a revised PBL formulation (Li and Arakawa 1997) have been conducted to investigate the sensitivity of the model climate to the "1CT" and "2CT" cloud optical properties. Our results show that the distinction between optically thin and thick clouds as assigned in the "2CT" cloud optical properties produces a more zonally asymmetric distribution of net surface SW radiation (Figure 2a) which is closer to the indirect observations of Li and Leighton (1993). At the same time, the enhanced low level cloud absorption and higher transmission in "2CT" results in a reduced stratus incidence over land and particularly over Brasil by lifting the condensation level above the PBL top.

4.2. From two to continuous cloud types

Releasing the fixed values in cloud optical properties (1CT & 2CT) has the advantage of being able to simulate a continuous spectrum of optical thicknesses and the elegance of relating the cloud optical properties directly to the cloud water mass, effective particle size and potentially sub-grid scale distribution. But this requires the prediction of the 1% of the atmospheric water existing as condensate. Preliminary results from our prognostic cloud water model (Köhler et al. 1997) show further improvement in the net surface SW flux particularly in the stratus regions off the coasts of Peru and Angola and a much reduced flux in the southern mid-latitudes dominated by stratiform clouds (Figure 2b).

Figure 2. AGCM simulation of January net surface SW flux using cloudiness parameterizations 2CT and CCT respectively.

---

5. Conclusions

We have documented the improvement in surface SW fluxes achievable by upgrading the cloud properties in an AGCM gradually. As an example of an indirect impact of that improvement we pointed out the sensitivity of the simulation of stratus clouds over land to the SW absorptivity and transmissivity of stratus.

Acknowledgments: We like to thank A. Arakawa and C.R. Mechoso for their encouragement, support and sharing their insight on this study. J.A. Spahr helped with the simulations. This research was supported by NSF grants ATM96-30226 and ATM92-24863 and by NASA grant NAG5-2591.

References

Harshvardhan, D.A. Randall, T.G. Corsetti, andD.A. Dazlich, 1989:Earth radiation budget and cloudiness simulations with a general circulation model. J. Atmos. Sci., 46, 1922-1942.

Katayama, A. 1972: A simplified scheme for computing radiative transfer in the troposphere. Numerical Simulation of Weather and Climate, Tech. Rep. No. 6. Dept. Meteor., University of California, Los Angeles, 77pp.

Köhler, et al. 1997: Ice cloud formulation in climate modeling. Seventh Conference on Climate Variations, Long Beach, CA, Amer. Met. Soc. , 237-242.

Li, J.-J. F., and A. Arakawa, 1997: Improved simulation of PBL moist processes with the UCLA AGCM. Seventh Conference on Climate Variations, Long Beach, CA, Amer. Met. Soc., 35-40.

Li, Z., and H.G. Leighton, 1993: Global climatologies of solar radiation budgets at the surface and in the atmosphere from 5 years of ERBE data. J. Geophys. Res., 98, 4919-4930.

Liao, X., W.B. Rossow, and D. Rind, 1995: Comparison between SAGE II and ISCCP high-level clouds. J. Geophys. Res., 100, 1121-1135.

Ma, C.-C., C.R. Mechoso, A. Arakawa, and John D. Farrara, 1994: Sensitivity of a coupled ocean-atmosphere model to physical parameterizations. J. Climate, 7, 1883-1896.

Rossow, W.B., and R.A. Schiffer, 1991: ISCCP cloud data products. Bul. Amer. Meteor. Soc., 72, 2-20.

Yu, J.-Y., and C.R. Mechoso, 1997: Impact of cloud radiative effects on the simulation of tropical Pacific with the UCLA coupled GCM. in preparation.