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

Title: An Evaluation of Marine Boundary Layer Cloud Property Simulations in the Community Atmosphere Model Using Satellite Observations: Conventional Subgrid Parameterization versus CLUBB

Abstract

This paper presents a two-step evaluation of the marine boundary layer (MBL) cloud properties from two Community Atmospheric Model (version 5.3, CAM5) simulations, one based on the CAM5 standard parameterization schemes (CAM5-Base), and the other on the Cloud Layers Unified By Binormals (CLUBB) scheme (CAM5-CLUBB). In the first step, we compare the cloud properties directly from model outputs between the two simulations. We find that the CAM5-CLUBB run produces more MBL clouds in the tropical and subtropical large-scale descending regions. Moreover, the stratocumulus (Sc) to cumulus (Cu) cloud regime transition is much smoother in CAM5-CLUBB than in CAM5-Base. In addition, in CAM5-Base we find some grid cells with very small low cloud fraction (<20%) to have very high in-cloud water content (mixing ratio up to 400mg/kg). We find no such grid cells in the CAM5-CLUBB run. However, we also note that both simulations, especially CAM5-CLUBB, produce a significant amount of “empty” low cloud cells with significant cloud fraction (up to 70%) and near-zero in-cloud water content. In the second step, we use satellite observations from CERES, MODIS and CloudSat to evaluate the simulated MBL cloud properties by employing the COSP satellite simulators. We note that a feature of the COSP-MODISmore » simulator to mimic the minimum detection threshold of MODIS cloud masking removes much more low clouds from CAM5-CLUBB than it does from CAM5-Base. This leads to a surprising result — in the large-scale descending regions CAM5-CLUBB has a smaller COSP-MODIS cloud fraction and weaker shortwave cloud radiative forcing than CAM5-Base. A sensitivity study suggests that this is because CAM5-CLUBB suffers more from the above-mentioned “empty” clouds issue than CAM5-Base. The COSP-MODIS cloud droplet effective radius in CAM5-CLUBB shows a spatial increase from coastal St toward Cu, which is in qualitative agreement with MODIS observations. In contrast, COSP-MODIS cloud droplet effective radius in CAM5-Base almost remains a constant. In comparison with CloudSat observations, the histogram of the radar reflectivity from modeled MBL clouds is too narrow without a distinct separation between cloud and drizzle modes. Moreover, the probability of drizzle in both simulations is almost twice as high as the observation. Future studies are needed to understand the causes of these differences and their potential connection with the “empty” cloud issues in the model.« less

Authors:
 [1];  [2];  [3];  [3];  [4]
  1. Joint Center for Earth Systems Technology, University of Maryland, Baltimore County, Baltimore, Maryland
  2. Joint Center for Earth Systems Technology, and Physics Department, University of Maryland, Baltimore County, Baltimore, Maryland
  3. Atmospheric Sciences and Global Change Division, Pacific Northwest National Laboratory, Richland, Washington
  4. Institute for Climate and Global Change Research, and School of Atmospheric Sciences, Nanjing University, Nanjing, China
Publication Date:
Research Org.:
Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1430421
Report Number(s):
PNNL-SA-125816
Journal ID: ISSN 0894-8755; KP1703010
DOE Contract Number:
AC05-76RL01830
Resource Type:
Journal Article
Resource Relation:
Journal Name: Journal of Climate; Journal Volume: 31; Journal Issue: 6
Country of Publication:
United States
Language:
English

Citation Formats

Song, Hua, Zhang, Zhibo, Ma, Po-Lun, Ghan, Steven J., and Wang, Minghuai. An Evaluation of Marine Boundary Layer Cloud Property Simulations in the Community Atmosphere Model Using Satellite Observations: Conventional Subgrid Parameterization versus CLUBB. United States: N. p., 2018. Web. doi:10.1175/JCLI-D-17-0277.1.
Song, Hua, Zhang, Zhibo, Ma, Po-Lun, Ghan, Steven J., & Wang, Minghuai. An Evaluation of Marine Boundary Layer Cloud Property Simulations in the Community Atmosphere Model Using Satellite Observations: Conventional Subgrid Parameterization versus CLUBB. United States. doi:10.1175/JCLI-D-17-0277.1.
Song, Hua, Zhang, Zhibo, Ma, Po-Lun, Ghan, Steven J., and Wang, Minghuai. Thu . "An Evaluation of Marine Boundary Layer Cloud Property Simulations in the Community Atmosphere Model Using Satellite Observations: Conventional Subgrid Parameterization versus CLUBB". United States. doi:10.1175/JCLI-D-17-0277.1.
@article{osti_1430421,
title = {An Evaluation of Marine Boundary Layer Cloud Property Simulations in the Community Atmosphere Model Using Satellite Observations: Conventional Subgrid Parameterization versus CLUBB},
author = {Song, Hua and Zhang, Zhibo and Ma, Po-Lun and Ghan, Steven J. and Wang, Minghuai},
abstractNote = {This paper presents a two-step evaluation of the marine boundary layer (MBL) cloud properties from two Community Atmospheric Model (version 5.3, CAM5) simulations, one based on the CAM5 standard parameterization schemes (CAM5-Base), and the other on the Cloud Layers Unified By Binormals (CLUBB) scheme (CAM5-CLUBB). In the first step, we compare the cloud properties directly from model outputs between the two simulations. We find that the CAM5-CLUBB run produces more MBL clouds in the tropical and subtropical large-scale descending regions. Moreover, the stratocumulus (Sc) to cumulus (Cu) cloud regime transition is much smoother in CAM5-CLUBB than in CAM5-Base. In addition, in CAM5-Base we find some grid cells with very small low cloud fraction (<20%) to have very high in-cloud water content (mixing ratio up to 400mg/kg). We find no such grid cells in the CAM5-CLUBB run. However, we also note that both simulations, especially CAM5-CLUBB, produce a significant amount of “empty” low cloud cells with significant cloud fraction (up to 70%) and near-zero in-cloud water content. In the second step, we use satellite observations from CERES, MODIS and CloudSat to evaluate the simulated MBL cloud properties by employing the COSP satellite simulators. We note that a feature of the COSP-MODIS simulator to mimic the minimum detection threshold of MODIS cloud masking removes much more low clouds from CAM5-CLUBB than it does from CAM5-Base. This leads to a surprising result — in the large-scale descending regions CAM5-CLUBB has a smaller COSP-MODIS cloud fraction and weaker shortwave cloud radiative forcing than CAM5-Base. A sensitivity study suggests that this is because CAM5-CLUBB suffers more from the above-mentioned “empty” clouds issue than CAM5-Base. The COSP-MODIS cloud droplet effective radius in CAM5-CLUBB shows a spatial increase from coastal St toward Cu, which is in qualitative agreement with MODIS observations. In contrast, COSP-MODIS cloud droplet effective radius in CAM5-Base almost remains a constant. In comparison with CloudSat observations, the histogram of the radar reflectivity from modeled MBL clouds is too narrow without a distinct separation between cloud and drizzle modes. Moreover, the probability of drizzle in both simulations is almost twice as high as the observation. Future studies are needed to understand the causes of these differences and their potential connection with the “empty” cloud issues in the model.},
doi = {10.1175/JCLI-D-17-0277.1},
journal = {Journal of Climate},
number = 6,
volume = 31,
place = {United States},
year = {Thu Mar 01 00:00:00 EST 2018},
month = {Thu Mar 01 00:00:00 EST 2018}
}