CCP is an interdivisional program, established 1 October 1994, to improve the basic knowledge of cloud-related processes and their parameterization in climate models. It is scientifically aligned with the WCRP (World Climate Research Programme) GCSS [GEWEX (Global Energy and Water-Cycle Experiment) Cloud System Study], which is discussed in another section of this report. CCP is co-led by Jeffrey Kiehl, of NCAR's Climate and Global Dynamics Division (CGD), and MMM's Mitchell Moncrieff.
While CCP started with MMM's basic research into cloud systems, activity is accelerating regarding the use of cloud resolving model (CRM) results to evaluate and improve convective parameterization. CRMs, which have 3D domains larger than the volumes studied via climate model grids, are CCP's research tool of choice. To this end, single-column versions of the NCAR Community Climate Model, Version 2 (CCM2) and Community Climate Model, Version 3 (CCM3) are being compared to domain-averaged properties obtained from the CRMs. Systematic studies are being set up as collaborative activities between MMM and CGD. CCP is thus moving from its initial phase, when most of the activity was at MMM, to one emphasizing interaction with other NCAR divisions.
Other examples of interdivisional-CCP collaboration are seen in additional efforts by CGD, Research Applications Program (RAP), and Atmospheric Technology Division (ATD). In CGD's Climate Modeling section, results from CRMs will be used to evaluate and improve convective parameterization methods. Moreover, 3D CRM results will be used as synthetic data in a scatterometer project within CGD's Oceanography Section, as well as by the Statistics Project in their Climate Analysis section. A satellite data analysis facility, reported last year by MMM, was put into operation by David Johnson during his two years of support by CCP. This facility is now a joint resource shared by RAP, ATD and MMM. It will be used in future CCP studies when a comprehensive analysis of satellite data is required.
The present CCP focus is on tropical precipitating convective cloud systems, due to their key climatic importance. To date, precedence has been given to cloud systems studied in GATE [GARP (Global Atmospheric Research Programme) Atlantic Tropical Experiment] and TOGA COARE (Tropical Ocean Global Atmosphere Coupled Ocean-Atmosphere Response Experiment), because these major field experiments are compatible with the cloud-resolving approach. That is, large-scale observations are used as diagnostic forcing, while the cloud system structure is explicitly modeled. Following simulations, the model data are evaluated against the observational data sets. Progress in this regard is reported in other sections of this report.
The conduct of workshops is another key activity of CCP, for the most part in conjunction with the GCSS Working Group on Precipitating Convective Cloud Systems (chaired by Moncrieff). For example, a workshop on the "Evaluation and Intercomparison of Cloud-Resolving Models Using TOGA COARE Observations" will be conducted in Annapolis, Maryland, 21-23 October 1996. A follow-up workshop is slated for Boulder in May 1997. Also, CCP will support participation in the GCSS/ECMWF (European Centre for Medium-Range Weather Forecasts) Workshop, "New Insights and Approaches in Convective Parameterization," in Reading, U.K., 4-7 November 1996.
A new CCP project, still in a fledgling state, is the study of convective systems in mid- to high-latitude cold air outbreaks. These cloud systems are also climatically important, but relatively little research has been conducted on them.
We finally mention that CRM data from CCP are being used by Graeme Stephens (Colorado State University), who is leading a major research initiative in which the properties of various kinds of cloud systems will be quantified by using active and passive remote sensing. It is anticipated that CRMs will be increasingly used in this capacity.
Chin-Hoh Moeng continued to work with the GEWEX Cloud System Study (GCSS) Boundary-Layer Cloud Working Group to examine the accuracy of various modeling applications for the stratocumulus-topped planetary boundary layer (PBL). One study focused on the applicability of large eddy simulation (LES). A second intercomparison study focused on a smoke-cloud case (no phase changes) to study entrainment. It is now being summarized for publication. A third intercomparison study concerns an Atlantic Stratocumulus Transition Experiment (ASTEX) drizzle case, designed to examine the effect of drizzle on turbulence dynamics within the tratocumulus cloud. This case was simulated, compared, and discussed during the International Cloud Modeling Workshop which took place in France in August 1996.
Changhai Liu (visitor, Texas A & M University) and Moncrieff investigated the diurnal variation of tropical oceanic convection through a series of idealized 2D cloud-resolving numerical experiments. Each experiment started from a horizontally homogeneous atmosphere upon which a uniform and time-independent large-scale forcing was imposed. The underlying surface was assumed to be an open ocean with a constant sea surface temperature. A pronounced diurnal cycle was simulated for organized convective systems; convective activity and intensity attained a maximum during the night and early morning and a minimum in the late afternoon. A similar but relatively weak diurnal variability was simulated for scattered or less organized convection. The modeled diurnal variation was primarily attributed to the direct interaction of radiation and convection, whereas the cloud-free differential heating mechanism played a secondary role. The cloud radiative interaction had a negative influence on convective activity, in contrast with general circulation modeling results. A full microphysics parameterization was found necessary to generate the diurnal cycle in the numerical experiments.
Wojciech Grabowski, Xiaoqing Wu (visitor, University of California, Los Angeles) and Moncrieff continued to study interactions of cloud dynamics and cloud microphysics with radiation and surface processes using GATE sounding data. Experiments using observed conditions simulated the evolution of cloud systems in the period 1-7 September 1974, during Phase III of GATE. The anelastic CRM was driven by the time-evolving large-scale horizontal wind field and time-evolving large-scale advective tendencies of temperature and moisture derived from observational budget-type studies. The seven-day period selected for the GATE case was characterized by the distinct evolution of cloud systems as a response to changing large-scale advective tendencies and low-level wind shear associated with easterly wave activity. Direct comparison was conducted between a variety of 2D simulations and a single 3D experiment which was a major computational effort. Three-dimensional results showed evolution of cloud systems with impressive similarities to the convection organization shown in the radar data. The 2D results were surprisingly similar to the 3D setup as far as mean thermodynamic fields are concerned. Work is underway to further quantify this conclusion.
Wu, Grabowski, and Moncrieff used available satellite, sounding, and surface data to evaluate model performance in long-term simulations of TOGA COARE cloud systems. The period selected included a major westerly wind burst observed during TOGA COARE and covered the period between 1 December 1992 and 10 January 1993. The period was characterized by the dramatic variability of cloud systems in response to changes of the large-scale forcing, horizontal winds, and ocean temperature. Detailed comparisons between model-generated fields and available data show very encouraging agreement as far as both short-term (few days) and long-term (few weeks) variability of model-produced cloud fields are concerned. This is very encouraging considering a rather poor spatial resolution and the two-dimensional framework applied in the simulations. The comparison prompted several modifications of the model physics (e.g., inclusion of the more sophisticated surface flux formulation). It also demonstrated a need for coupling the cloud model with the ocean model to further quantify effects of cloud-scale processes (e.g., radiative cloud forcing, precipitation) on the upper ocean dynamics and heat budget over the tropical western Pacific.
To complement last year's numerical studies, Liu and Moncrieff developed a nonlinear analytic model to study the bulk characteristics of energy-conserving density currents in stratified and sheared environments. The idealized representation of latent heating in a stratified flow is a unique feature that interactively couples the dynamic and thermodynamic fields. A stable stratification decreases the height of density currents but increases the corresponding propagation speed. In contrast, the density current is deeper and moves more slowly once latent heating is included. As for the effect of shear, the depth and translation speed of density currents increase as the ambient shear varies from negative to positive (in the direction of propagation), with the exception of a strongly stable environment. A key addition to density current dynamics is the upper-level overturning circulation ahead of the system. This feature is very different from the blocked or choked upper-level structure found in the companion study by Liu and Moncrieff (1996). The distinction is attributed to the effect of different shear profiles on density current dynamics.
The analytic results quantifying the roles of shear and latent heating in density current-like phenomena in the atmosphere needs to be evaluated against high-resolution numerical simulations and observations, such as those made during the Maritime Continent Thunderstorm Experiment (MCTEX).
Jun-Ichi Yano (Climate Research Centre, Monash University, Australia) and Moncrieff completed a linear model of the impact of convective momentum transport on the large-scale circulation of the tropics. This was done by integrating an archetypal model for mesoscale momentum flux, developed by Moncrieff in 1992, into a linear model previously used by Yano and Emanuel (1991) to investigate the Wind-Induced Surface Heat Experiment (WISHE) instability. In a one-layer representation, the WISHE instability is suppressed by the mesoscale momentum flux because the larger induced effective stratification provides a more efficient mass redistribution, which further provides a larger potential energy for WISHE. In a two-layer model, the modulation of the mesoscale momentum flux by convective activity is also taken into account. The WISHE instability exists in most cases, albeit generally weakened by the mesoscale momentum flux, and the phase speed of the unstable disturbance is reduced. Finally, a new type of the instability is discovered, which is induced by a self-enhancement of the circulation by the mesoscale momentum transfer coupled to gravity-wave radiative damping at the top of the atmosphere. This propagates much slower than the WISHE mode, and turns from traveling up-surface-wind to down-surface-wind when the barotropic mean flow is sufficiently strong. This instability can cause a self-maintaining mechanism for mesoscale organization.
Based on environmental conditions of the 22 and 23 June squall lines during the Convection Profonde Tropicale in 1981 (COPT81) experiment in West Africa, Liu, Moncrieff and Edward Zipser (Texas A & M University) conducted a series of 2D numerical simulations to examine the effect of microphysics on tropical squall line dynamics, including the various effects of ice physics. The effect of ice-phase microphysics was found to depend strongly on ambient conditions. For the environment of strong convective instability, ice impact is important on the system-scale structure but is not essential to the convective-scale dynamics. On the other hand, the ice influence is crucial to producing the squall-line convective system if the environment has a weak convective instability and is almost saturated at low levels. We concluded that the terminal velocity effect was important to realistically simulate the stratiform structure. However, the melting process was responsible for the formation of the subcloud cold pool in a weakly unstable and almost saturated environment.
Numerical simulations of a fast-moving tropical squall line and slow-moving banded tropical convection by William Skamarock and Stanley Trier were analyzed by Trier and Margaret LeMone to determine their impact on the PBL. For the squall line case, the thermodynamic structure of the boundary layer was found to be influenced foremost by the origination level of convective and mesoscale downdrafts. However, there is also spatial variability in the boundary layer thermodynamic structure that is related both to the highly variable amount of rainfall evaporation along downdraft trajectories, and to the local strength of oceanic heat fluxes. The slow-moving banded convection was simulated using the environmental temperature and moisture structure of the squall case, but with a wind profile obtained in the environment of the 9 February 1993 banded mesoscale convection system which was observed and analyzed by Bradley Smull and Thomas Matejka (both of National Severe Storms Laboratory, National Oceanic and Atmospheric Administration), and LeMone. Unlike the fast-moving squall case, whose motion was continuous, the simulated slow-moving banded convection propagated (as in the observations) by discrete development of narrow precipitation bands. The heat and freshwater fluxes (rain) were somewhat weaker and much more localized with the slow-moving banded convection than in the fast-moving squall line case, despite similar environmental surface wind speeds.
The data from the above-described tropical simulations were used, by Trier, LeMone, and Skamarock, in collaboration with Steven Anderson and Robert Weller (both of the Woods Hole Oceanographic Institute), to study the effects of mesoscale precipitation systems on the structure of the ocean mixed layer. In this study, a mixed layer model was forced with cloud-model surface heat (latent and sensible) fluxes, stress, and freshwater fluxes. The ocean model results showed an evolution of the ocean mixed layer after precipitation events that is similar to observations taken during TOGA COARE. It is clear from the simulation that oceanic convection, which results from buoyancy fluxes associated with the precipitation events and the diurnal cycle of shortwave radiation, must be considered in addition to wind-driven mixing to explain the observed evolution of the ocean mixed layer. The current simulation and planned sensitivity studies are expected to add to the current understanding of mechanisms for local development of salt-stratified barrier layers in the western equatorial Pacific.
As an extension of this work, Zipser decided to take a closer look at the balloon soundings in the COARE Intensive Flux Array (IFA) to see what information they could add to the study. He and his graduate student, Chris Lucas, found that cluster analysis successfully isolated the similar convection environmental soundings, as well as soundings of other types (Lucas and Zipser, 1996). They also found, however, (a) a systematic variation of CAPE values among ships in the COARE flux array, and (b) distinct types of soundings that seemed to sort out by ship. In view of relatively uniform sea surface temperatures and surface conditions (available as the lowest level for the soundings), and expected surface-layer/mixed-layer differences from Monin-Obukhov theory, they detected biases in mixing ratio of 1-2 gm/kg at some ships, with comparable scatter in the data. Similar behavior was confirmed by David Parsons (ATD) for some island stations that used the same Vaisala sonde. By applying corrections for the mean biases and forcing the surface-/mixed-layer differences to go halfway to matching the typical values of 1.25 gm/kg, they obtained a dataset which (a) allowed sounding types to be more evenly distributed among the IFA ships, and (b) produced average CAPE profiles that were more uniform (30 per cent difference) than previously (factor of three difference). Zipser, Lucas, and LeMone are now working with Paul Ciesielski and Richard Johnson (both of Colorado State University), Parsons, and several other COARE scientists to find out how general the problem is (do the other sondes have similar problems?), and to look for the best method of mitigating the data problem in an improved sounding data set.
Jielun Sun (visitor, University of Colorado, Boulder) in collaboration with LeMone, Larry Mahrt (Affiliate Scientist, Oregon State University) and Robert Grossman (University of Colorado) analyzed the horizontal scale dependency of vertical eddy fluxes in the tropical marine boundary layer using the NCAR Electra aircraft data from TOGA COARE. They studied the scale dependence of flux, and how it relates to the bulk aerodynamic formulation and parameterization of subgrid-scale flux. The results showed that fluxes are sensitive to the choice of cut-off wavelength in the presence of significant mesoscale activity, and in weak wind cases where the turbulent fluxes are small. Mesoscale heat, moisture and momentum fluxes for individual flight legs can reach 20% of the turbulent fluxes in the presence of well-organized, convective cloud systems, even as low as 35 m above the sea surface. The mesoscale flux is less correlated to the wind speed and bulk air-sea difference than turbulent fluxes. The local mesoscale flux can be upward or downward, and therefore its average value decreases as the averaging length increases. The mesoscale momentum flux is less systematic than the turbulent stress and is more sensitive to flux averaging scale than the turbulent stress. Results of this study have been submitted to the Journal of the Atmospheric Sciences.