Atmospheric Science [A]

 CC:Hall E  Monday  1400h

Tropical Convection: Observations, Theory, High-Resolution Modeling, and Parameterization IV Posters

Presiding:  G Zhang, Scripps Institution of Oceanography; M Zhang, Stony Brook University


Temporal Relations of Column Water Vapor and Precipitation

* Holloway, C E (, University of Reading and NCAS-Climate, Department of Meteorology University of Reading Earley Gate, PO Box 243, Reading, RG6 6BB, United Kingdom
Neelin, J (, UCLA, Dept. of Atmospheric and Oceanic Sciences University of California, Los Angeles 405 Hilgard Ave., Los Angeles, CA 90095, United States

Empirical studies using satellite data and radiosondes have shown that precipitation increases with column water vapor (CWV) in the tropics, and that this increase is much steeper above some critical CWV value. Here, eight years of 1-minute resolution microwave radiometer and optical gauge data at the Atmospheric Radiation Measurement (ARM) site on Nauru Island are analyzed to better understand the relationships between CWV, column liquid water (CLW), cloud top height, and precipitation at small time scales. CWV is found to have large autocorrelation times compared with CLW and precipitation. Before precipitation events, CWV increases on both a synoptic-scale time period and a subsequent shorter time period consistent with mesoscale convective activity---the latter period is associated with the highest CWV levels. Probabilities of precipitation increase greatly with CWV: 10--12 hr after high CWV, there are still significantly higher probabilities. Even in periods of high CWV, probabilities of initial precipitation in a 5-minute period remain low enough that there tends to be a lag before the start of the next precipitation event. This suggests that CWV can be a useful predictor for precipitation in stochastic convective parameterizations.


Simulations of the Precipitation Using WRF and Comparisons with Satellite Observations and CAM

* Murthi, A (, Texas A&M University, 3150 TAMU, College Station, TX 77843, United States
Bowman, K (, Texas A&M University, 3150 TAMU, College Station, TX 77843, United States
Leung, R (, Pacific Northwest National Laboratory, PO Box 999 MSIN: K9-24, Richland, WA 99352, United States

The accurate representation of rainfall in models of global climate has been a challenging task for climate modelers owing to its small space and time scales. Quantifying this variability is important for comparing simulations of atmospheric behavior with real time observations. In this regard, this paper compares the statistical aspects of rainfall simulated by the high- resolution (36 km) Weather Research Forecast (WRF) model with satellite observations from the Tropical Rainfall Measuring Mission (TRMM) and simulations from the Community Atmosphere Model (CAM) at variable spatial resolutions T85 and T42. Six years worth of rainfall rate data (2000 - 2005) from within the Tropics (30°S -- 30°N) have been used in the analysis and results are presented in terms of long-term mean rain rates, probability density functions (PDFs) and amplitude and phase of the diurnal cycle.
The results indicate that the WRF model agrees better with the observations than CAM. This is most evident in the spatial distributions of the long-term means and in the PDFs of rain rate wherein WRF seems to simulate the heavy rain rates better than CAM. One of the primary goals of this study is to identify physical mechanisms that are responsible for the "propagating" rainfall diurnal cycle and its corresponding representation in WRF. Preliminary analysis of the results indicates that WRF simulates the spatial pattern of the amplitudes of the diurnal cycle fairly well as compared to CAM which over/under approximates the amplitudes almost everywhere in the Tropics. The phase of the diurnal cycle as simulated by WRF displays off-shore phase propagation along the coasts of some of the major islands in the Maritime continent in accordance with the observations. This feature is absent in CAM which generates a nearly uniform phase signal over both land and ocean regions. Thus these results suggest that the 36-km WRF model is able to simulate both the forced ("diurnal cycle") and the stochastic ("PDFs") component of rain fall variability better than CAM.


Relationship between Cloud Base Aerosols and Cloud Droplet Concentrations Derived from Three Years of Airborne Measurements in West Africa

* Delene, D J (, University of North Dakota, Department of Atmospheric Sciences 4149 University Avenue, Grand Forks, ND 58202-9006, United States

Aircraft measurements of below cloud base aerosols and above cloud base droplets are the most direct method to quantify the effect aerosols have on cloud microphysical properties. Understanding how aerosols affect cloud microphysics is an important step in understanding the precipitation formation process. Airborne measurements have been made during the wet season in Mali, West Africa during 2006, 2007, and 2008. Below cloud base aerosols properties were sampled using a Passive Cavity Aerosol Spectrometer (PCASP) and a Cloud Condensation Nuclei Counter (CCNC). The PCASP measured the aerosol size spectrum between 0.1 and 3.0 µm, while the CCNC measured the number concentration of aerosols that activate at a prescribed supersaturation. The above cloud base droplet spectrum between 3.0 and 50 µm was measured by a Forward Scattering Spectrometer Probe (FSSP). The relationship between cloud base total PCASP aerosol concentrations, PCASP aerosol median diameter, CCN concentration at 1% supersaturation, FSSP total droplet concentration, and FSSP median volume diameter were computed for each year's measurements. The relationships show marked differences among the three years of airborne measurements. Future work will focus on extending the airborne measurements to longer time periods by using satellite derived aerosol optical depth and radar derived cloud properties.


Statistical Characterization of Simulated Cumulus Convection: Toward Improving Stochastic Convective Parameterizations

* Jones, T R (, Department of Atmospheric Science, Colorado State University, 1371 Atmospheric Science, Fort Collins, CO 80525,
Randall, D A (, Department of Atmospheric Science, Colorado State University, 1371 Atmospheric Science, Fort Collins, CO 80525,

The desire to create stochastic convective parameterizations (SCPs) has developed from the realization that quasi-equilibrium (QE)-based convective parameterizations fail to reproduce the full spectrum of convective variability, when employed in global circulation models (GCMs) with low spatial resolution, that is found in CRM ensembles and observational data. Implementation of SCPs can be as simple as introducing a random multiplier to variables in a given parameterization to increase overall ensemble spread and improve probabilistic precipitation forecasts, but such an approach is not a true physical parameterization, directly linked to resolved processes. A more complex, yet physically based, method requires an understanding of the nature of the deviation from QE to be able to direct convective variability in a more informed manner. To obtain a characterization of 'true' convective variability, the three-dimensional Jung-Arakawa anelastic cloud- resolving model, which uses the vector vorticity equation in its dynamical core, is used in this study with a 2-km horizontal resolution in place of an observational dataset. A number of simulations are performed to study the non-equilibrium, stochastic component of moist convective heating and drying. Following Xu et al. (1992), the response of the non-deterministic component of the numerical simulations is tested by means of cyclic prescribed forcings with periods ranging from 2 to 120 hours. Additionally, the dependence of the simulation characteristics on the size of the computational domain is investigated. It is found that the variability around QE is itself variable with dependencies on multiple parameters. It varies most strongly with changes in the size of the computational domain (i.e. grid size in a GCM) with a large shift in variability about QE when moving from 256 km to 128 km grid spacing. This finding is of particular importance as grid spacing in GCM is trending toward this point, downward from (O)1000 km. Stochastically adding in this previously lacking variability is a task which will need to, therefore, take into account a number of parameters, including the chosen grid spacing and the length scale of the large-scale forcing.