skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: COLLABORATIVE RESEARCH:USING ARM OBSERVATIONS & ADVANCED STATISTICAL TECHNIQUES TO EVALUATE CAM3 CLOUDS FOR DEVELOPMENT OF STOCHASTIC CLOUD-RADIATION

Abstract

The long-range goal of several past and current projects in our DOE-supported research has been the development of new and improved parameterizations of cloud-radiation effects and related processes, using ARM data, and the implementation and testing of these parameterizations in global models. The main objective of the present project being reported on here has been to develop and apply advanced statistical techniques, including Bayesian posterior estimates, to diagnose and evaluate features of both observed and simulated clouds. The research carried out under this project has been novel in two important ways. The first is that it is a key step in the development of practical stochastic cloud-radiation parameterizations, a new category of parameterizations that offers great promise for overcoming many shortcomings of conventional schemes. The second is that this work has brought powerful new tools to bear on the problem, because it has been a collaboration between a meteorologist with long experience in ARM research (Somerville) and a mathematician who is an expert on a class of advanced statistical techniques that are well-suited for diagnosing model cloud simulations using ARM observations (Shen).

Authors:
Publication Date:
Research Org.:
The Regents of the University of California / Scripps Institution of Oceanography
Sponsoring Org.:
USDOE
OSTI Identifier:
1091953
Report Number(s):
Final Report
DOE Contract Number:
SC0000658
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
15 GEOTHERMAL ENERGY; cloud, radiation, statistical, ARM, observations

Citation Formats

Somerville, Richard. COLLABORATIVE RESEARCH:USING ARM OBSERVATIONS & ADVANCED STATISTICAL TECHNIQUES TO EVALUATE CAM3 CLOUDS FOR DEVELOPMENT OF STOCHASTIC CLOUD-RADIATION. United States: N. p., 2013. Web. doi:10.2172/1091953.
Copy to clipboard
Somerville, Richard. COLLABORATIVE RESEARCH:USING ARM OBSERVATIONS & ADVANCED STATISTICAL TECHNIQUES TO EVALUATE CAM3 CLOUDS FOR DEVELOPMENT OF STOCHASTIC CLOUD-RADIATION. United States. doi:10.2172/1091953.
Copy to clipboard
Somerville, Richard. Thu . "COLLABORATIVE RESEARCH:USING ARM OBSERVATIONS & ADVANCED STATISTICAL TECHNIQUES TO EVALUATE CAM3 CLOUDS FOR DEVELOPMENT OF STOCHASTIC CLOUD-RADIATION". United States. doi:10.2172/1091953. https://www.osti.gov/servlets/purl/1091953.
Copy to clipboard
@article{osti_1091953,
title = {COLLABORATIVE RESEARCH:USING ARM OBSERVATIONS & ADVANCED STATISTICAL TECHNIQUES TO EVALUATE CAM3 CLOUDS FOR DEVELOPMENT OF STOCHASTIC CLOUD-RADIATION},
author = {Somerville, Richard},
abstractNote = {The long-range goal of several past and current projects in our DOE-supported research has been the development of new and improved parameterizations of cloud-radiation effects and related processes, using ARM data, and the implementation and testing of these parameterizations in global models. The main objective of the present project being reported on here has been to develop and apply advanced statistical techniques, including Bayesian posterior estimates, to diagnose and evaluate features of both observed and simulated clouds. The research carried out under this project has been novel in two important ways. The first is that it is a key step in the development of practical stochastic cloud-radiation parameterizations, a new category of parameterizations that offers great promise for overcoming many shortcomings of conventional schemes. The second is that this work has brought powerful new tools to bear on the problem, because it has been a collaboration between a meteorologist with long experience in ARM research (Somerville) and a mathematician who is an expert on a class of advanced statistical techniques that are well-suited for diagnosing model cloud simulations using ARM observations (Shen).},
doi = {10.2172/1091953},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Thu Aug 22 00:00:00 EDT 2013},
month = {Thu Aug 22 00:00:00 EDT 2013}
}
Copy to clipboard
Technical Report:
View Technical Report (3.22 MB)
DOI:  10.2172/1091953
Save / Share:
Export Metadata
Save to My Library
You must Sign In or Create an Account in order to save documents to your library.

  • The long-range goal of several past and current projects in our DOE-supported research has been the development of new and improved parameterizations of cloud-radiation effects and related processes, using ARM data, and the implementation and testing of these parameterizations in global models. The main objective of the present project being reported on here has been to develop and apply advanced statistical techniques, including Bayesian posterior estimates, to diagnose and evaluate features of both observed and simulated clouds. The research carried out under this project has been novel in two important ways. The first is that it is a key stepmore » in the development of practical stochastic cloud-radiation parameterizations, a new category of parameterizations that offers great promise for overcoming many shortcomings of conventional schemes. The second is that this work has brought powerful new tools to bear on the problem, because it has been an interdisciplinary collaboration between a meteorologist with long experience in ARM research (Somerville) and a mathematician who is an expert on a class of advanced statistical techniques that are well-suited for diagnosing model cloud simulations using ARM observations (Shen). The motivation and long-term goal underlying this work is the utilization of stochastic radiative transfer theory (Lane-Veron and Somerville, 2004; Lane et al., 2002) to develop a new class of parametric representations of cloud-radiation interactions and closely related processes for atmospheric models. The theoretical advantage of the stochastic approach is that it can accurately calculate the radiative heating rates through a broken cloud layer without requiring an exact description of the cloud geometry.« less

  • ; ; ; ...
    This report briefly summarizes the progress made by ARM postdoctoral fellow, Yanluan Lin, at GFDL during the period from October 2008 to present. Several ARM datasets have been used for GFDL model evaluation, understanding, and improvement. This includes a new ice fall speed parameterization with riming impact and its test in GFDL AM3, evaluation of model cloud and radiation diurnal and seasonal variation using ARM CMBE data, model ice water content evaluation using ARM cirrus data, and coordination of the TWPICE global model intercomparison. The work illustrates the potential and importance of ARM data for GCM evaluation, understanding, and ultimately,more » improvement of GCM cloud and radiation parameterizations. Future work includes evaluation and improvement of the new dynamicsPDF cloud scheme and aerosol activation in the GFDL model.« less

  • ; ; ; ...
    Mixed-phase clouds are composed of a mixture of cloud droplets and ice crystals. The cloud microphysics in mixed-phase clouds can significantly impact cloud optical depth, cloud radiative forcing, and cloud coverage. However, the treatment of mixed-phase clouds in most current climate models is crude and the partitioning of condensed water into liquid droplets and ice crystals is prescribed as temperature dependent functions. In our previous 2007 ARM metric reports a new mixed-phase cloud microphysics parameterization (for ice nucleation and water vapor deposition) was documented and implemented in the NCAR Community Atmospheric Model Version 3 (CAM3). The new scheme was testedmore » against the Atmospheric Radiation Measurement (ARM) Mixed-phase Arctic Cloud Experiment (M-PACE) observations using the single column modeling and short-range weather forecast approaches. In this report this new parameterization is further tested with CAM3 in its climate simulations. It is shown that the predicted ice water content from CAM3 with the new parameterization is in better agreement with the ARM measurements at the Southern Great Plain (SGP) site for the mixed-phase clouds.« less

  • ; ;
    GFDL CM3 convective vertical velocities, parameterized because their scales are below those resolved in current climate models, have been compared with observations from DOE ARM. Single-column forcing has been used for this comparison for two time periods from TWP-ICE and three from MC3E. These are the first independent evaluations of the parameterized vertical velocities against observations. The results show that basic characteristics of the observations are captured by CM3 when forced with single-column observations and constrained to observed large-scale states. In many, but not all, cases, the parameterized vertical velocities exceed those observed, especially at higher levels in the clouds.more » Similar problems have been noted by others in cloud-resolving models. The results have important implications for cloud-aerosol interactions, which depend on vertical velocities, and likely pose constraints on entrainment by convection, which may be related to climate sensitivity.« less

  • The representation of clouds in Global Climate Models (GCMs) remains a major source of uncertainty in climate change simulations. Cloud climatologies have been widely used to either evaluate climate model cloud fields or examine, in combination with other data sets, climate-scale relationships between cloud properties and dynamical or microphysical parameters. Major cloud climatologies have been based either on satellite observations of cloud properties or on surface observers views of cloud type and amount. Such data sets provide either the top-down view of column-integrated cloud properties (satellites) or the bottom-up view of the cloud field morphology (surface observers). Both satellite-based andmore » surface cloud climatologies have been successfully used to examine cloud properties, to support process studies, and to evaluate climate and weather models. However, they also present certain limitations, since the satellite cloud types are defined using radiative cloud boundaries and surface observations are based on cloud boundaries visible to human observers. As a result, these data sets do not resolve the vertical distribution of cloud layers, an issue that is important in calculating both the radiative and the hydrologic effects of the cloud field. Ground-based cloud radar observations, on the other hand, resolve with good accuracy the vertical distribution of cloud layers and could be used to produce cloud type climatologies with vertical layering information. However, these observations provide point measurements only and it is not immediately clear to what extent they are representative of larger regimes. There are different methods that can be applied to minimize this problem and to produce cloud layering climatologies useful for both cloud process and model evaluation studies. If a radar system is run continuously over a number of years, it eventually samples a large number of dynamical and microphysical regimes. If additional data sets are used to put the cloud layering information into the context of large-scale dynamical regimes, such information can be used to study interactions among cloud vertical distributions and dynamical and microphysical processes and to evaluate the ability of models to simulate those interactions. The U.S. Department of Energy's Atmospheric Radiation Measurement (ARM) program has established several Climate Research Facilities (ACRF) that provide continuous, long-term observations of clouds and radiation. ARM, with its overall goal of improving the treatment of radiation and clouds in climate models has provided unique observing systems for accelerating progress on the representation of cloud processes. In this project, six and a half years (January 1998 to June 2004) of cloud observations collected at the Southern Great Plains (SGP) Oklahoma ACRF were used to produce a cloud-type climatology. The climatology provides cloud amounts for seven different cloud types as well as information on the detailed structure of multi-layer cloud occurrences. Furthermore, the European Centre for Medium-Range Weather Forecasts (ECMWF) model output was used to define the dynamic regimes present during the observations of the cloud conditions by the vertically pointing radars at the SGP ACRF. The cloud-type climatology and the ECMWF SGP data set were then analyzed to examine and map dynamical conditions that favor the creation of single-layer versus multi-layer cloud structures as well as dynamical conditions that favor the occurrence of drizzle in continental stratus clouds. In addition, output from the ECMWF weather model forecasts was analyzed with the objective to compare model and radar derived cloud type statistics, in order to identify the major model deficiencies in cloud vertical distribution and map their seasonal variations. The project included two primary goals. The first was to create a cloud type climatology over the Southern Great Planes site that will show how cloud vertical distribution varies with dynamic and thermodynamic regime and how these variations would affect cloud climate feedbacks. The second was to compare this climatology to clouds derived by a numerical model in order to identify the major model deficiencies in cloud vertical distribution and map their seasonal variations.« less