A Bimodal Diagnostic Cloud Fraction Parameterization. Part II: Evaluation and Resolution Sensitivity

View More View Less
  • 1 Met Office, Exeter, United Kingdom
  • 2 Met Office, Exeter, United Kingdom
  • 3 Met Office, Exeter, United Kingdom
© Get Permissions
Restricted access

Abstract

A wide range of approaches exists to account for subgrid cloud variability in regional simulations of the atmosphere. This paper addresses the following questions: (1) Is there still benefit in representing subgrid variability of cloud in convection-permitting simulations? (2) What is the sensitivity to the cloud fraction parameterization complexity? (3) Are current cloud fraction parameterizations scale-aware across convection-permitting resolutions? These questions are addressed for regional simulations of a six-week observation campaign in the US Southern Great Plains. Particular attention is given to a new diagnostic cloud fraction scheme with a bimodal subgrid saturation-departure PDF, described in Part I. The model evaluation is performed using ground-based remote sensing synergies, satellite-based retrievals and surface observations. It is shown that not using a cloud-fraction parameterization results in underestimated cloud frequency and water content, even for stratocumulus. The use of a cloud-fraction parameterization does not guarantee improved cloud property simulations, however. Diagnostic and prognostic cloud schemes with a symmetric subgrid saturation-departure PDF underestimate cloud fraction and cloud optical thickness, and hence overestimate surface shortwave radiation. These schemes require empirical bias-correction techniques to improve the cloud cover. The new cloud-fraction parameterization, introduced in Part I, improves cloud cover, liquid water content, cloud base height, optical thickness and surface radiation compared to schemes reliant on a symmetric PDF. Furthermore, cloud parameterizations using turbulence-based, rather than prescribed constant subgrid variances, are shown to be more scale-aware across convection-permitting resolutions.

Corresponding author address: Atmospheric Processes and Parametrizations, Met Office, FitzRoy Road, Exeter, United Kingdom. E-mail: kwinten.vanweverberg@metoffice.gov.uk

Abstract

A wide range of approaches exists to account for subgrid cloud variability in regional simulations of the atmosphere. This paper addresses the following questions: (1) Is there still benefit in representing subgrid variability of cloud in convection-permitting simulations? (2) What is the sensitivity to the cloud fraction parameterization complexity? (3) Are current cloud fraction parameterizations scale-aware across convection-permitting resolutions? These questions are addressed for regional simulations of a six-week observation campaign in the US Southern Great Plains. Particular attention is given to a new diagnostic cloud fraction scheme with a bimodal subgrid saturation-departure PDF, described in Part I. The model evaluation is performed using ground-based remote sensing synergies, satellite-based retrievals and surface observations. It is shown that not using a cloud-fraction parameterization results in underestimated cloud frequency and water content, even for stratocumulus. The use of a cloud-fraction parameterization does not guarantee improved cloud property simulations, however. Diagnostic and prognostic cloud schemes with a symmetric subgrid saturation-departure PDF underestimate cloud fraction and cloud optical thickness, and hence overestimate surface shortwave radiation. These schemes require empirical bias-correction techniques to improve the cloud cover. The new cloud-fraction parameterization, introduced in Part I, improves cloud cover, liquid water content, cloud base height, optical thickness and surface radiation compared to schemes reliant on a symmetric PDF. Furthermore, cloud parameterizations using turbulence-based, rather than prescribed constant subgrid variances, are shown to be more scale-aware across convection-permitting resolutions.

Corresponding author address: Atmospheric Processes and Parametrizations, Met Office, FitzRoy Road, Exeter, United Kingdom. E-mail: kwinten.vanweverberg@metoffice.gov.uk
Save