Cloud Modeling

• Cirrus Modeling
• Overview of Regional Atmospheric Modeling System (RAMS)
• Sample results for the Idealized GCSS Test Case
• ICRCCM II - Cloud Model Fields

• Cirrus Modeling

Cirrus clouds are known to have an important impact on Earth Radiation Budget through their radiative effect and an influence on atmospheric motion on many time and space scales; but they also have been identified as a major uncertainty in diagnosing and forecasting climate change. Various numerical studies have indicated that the radiative effects of these cloud depend on their physical and mycrophysical properties and the net contribution to surface radiative cooling/heating can be quite different depending on the cloud properties.

From the point of view of dynamics, cirrus clouds are interesting convectively active systems. The dynamics of their formation and evolution need to be further investigated in order to obtain as much information as possible from the observations and develop more efficient retrieval schemes.

A good parameterization of cirrus clouds, and ice clouds in general, is also needed in weather forecast models: the way clouds are treated in large-scale models influence quantities that determine the heat and moisture budgets of the planets (see Stephens et al. in GEWEX NEWS, February 1998.

The need for more organized cloud modeling activities to address climate and weather issues has been recognized in the framework of the Global Energy and Water Cycle Experiment (GEWEX) and a special project, the GEWEX Cloud System Study (GCSS) was organized. The aim of the GCSS is to gather and coordinate the efforts of mesoscale cloud modelers and cloud observers to improve the present understanding of cloud properties and cloud-radiation interaction, to develop better cloud parameterizations for GCMs and weather forecast models. In particular, the efforts of GCSS Working Group 2 (WG-2) are aimed to study cirrus clouds both using Cloud Resolving Models (CMR) and observational data.

One of the first issue WG-2 addressed to, has been the intercomparison of different models. The participants were encouraged to run their models and simulate the same test cases in order to be able to distinguish between the model-induced features and the real cloud properties. The description of the idealized test case and the pressure, temperature and relative humidity profiles can be found in the WG-2 home page.

As part of this project the Regional Atmospheric Modeling System (RAMS) developed at the Department of Atmospheric Sciences of the Colorado State University was used to performed the idealized test case simulation.




---

• Overview of RAMS

The RAMS model merges a non--hydrostatic cloud model (Tripoli and Cotton, 1982) and a mesoscale model (Mahrer and Pielke, 1977). The main prognosed variables are the wind components, the ice--liquid water potential temperature, the perturbation Exner function, the total liquid water mass mixing ratio, the mixing ratios and the number concentrations of different hydrometeor species. From this variables, temperature, potential temperature, vapor mixing ratio and cloud--water mixing ratio are diagnosed.

The grid structures includes a polar--stereographic grid in the horizontal and a terrain--following vertical coordinate. Vertical grid spacing is user--defined. The top boundary condition is w=0 (rigid lid) and to avoid possible reflection a damping method to absorb wave energy is used. At the bottom, w=0, and the surface temperature is assumed constant. The lateral boundary conditions are periodic.

The cloud mycrophysics module predicts the mixing ratio of pristine ice crystals, aggregates, rain and cloud water. Nucleation, vapor deposition growth, collection and precipitation--settling processes are also simulated.

For the simulation performed only the significant hydrometeors, such as pristine ice (small ice crystals), snow (bigger ice crystal, threshold diameter D=125 micron) and aggregates, were included, whereas rain, graupel and hail were turned off, being less likely present in a high ice cloud. Cloud droplets, which are assumed to be supercoooled, and pristine ice are allowed to nucleate only from vapor, while all the other categories form from pre--existing hydrometeors and then grow by vapor deposition.

Nucleation results from two primary processes:

1. A combination of vapor deposition and condensation--freezing mechanism for which the amount of activated nuclei is derived from the amount of supersaturation with respect to ice;
2. Contact nucleation for which the number of potential nuclei is derived from an observed temperature dependence.

The radiation scheme used in this version of RAMS is a two--stream model with 3 bands in the shortwave and 5 in the longwave. The required inputs for the radiation routine are: optical depth, single scattering albedo and asymmetry parameter.

	After 15 min from the beginning of the simulation the cloud appears relatively stratiform and shows low values of total ice mixing ratio.
	At t=30 min into the simulation the cloud has deepened and seems more convectively active. Ice mixing ratio has increased up to an order of magnitude higher than in the previous plot, especially in the convective pockets.
	As the simulation progresses, the convective cells in the cloud seem to develop a symmetric structure. The ice mixing ratio remained almost constant. The cloud base is slowly lowering.
	The cloud structure appears to be very similar to the previous picture, except for the cloud base which is a little lower and the convective cells which appear elongated in the horizontal.
	At t=75 min, the cloud still shows the same structure but the maximum peak value of ice mixing ratio is decreasing.
	After 1 hour and 30 min into the simulation, the cloud seems to be dissipating (see left side of the panel). The convective cells are stil visible.
	Similarly at t=1 hour 45 min, the cloud is evaporating at its base and the ice mixing raio is further decreasing.