GTS v1.0: a macrophysics scheme for climate models based on a probability density function

Abstract

Abstract. Cloud macrophysics schemes are unique parameterizations for general circulation models. We propose an approach based on a probability density function (PDF) that utilizes cloud condensates and saturation ratios to replace the assumption of critical relative humidity (RH). We test this approach, called the Global Forecast System (GFS) – Taiwan Earth System Model (TaiESM) – Sundqvist (GTS) scheme, using the macrophysics scheme within the Community Atmosphere Model version 5.3 (CAM5.3) framework. Via single-column model results, the new approach simulates the cloud fraction (CF)–RH distributions closer to those of the observations when compared to those of the default CAM5.3 scheme. We also validate the impact of the GTS scheme on global climate simulations with satellite observations. The simulated CF is comparable to CloudSat/Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) data. Comparisons of the vertical distributions of CF and cloud water content (CWC), as functions of large-scale dynamic and thermodynamic parameters, with the CloudSat/CALIPSO data suggest that the GTS scheme can closely simulate observations. This is particularly noticeable for thermodynamic parameters, such as RH, upper-tropospheric temperature, and total precipitable water, implying that our scheme can simulate variation in CF associated with RH more reliably than the default scheme. Changes in CF and CWC would affect climatic fields and large-scale circulation via cloud–radiation interaction. Both climatological means and annual cycles of many of the GTS-simulated variables are improved compared with the default scheme, particularly with respect to water vapor and RH fields. Different PDF shapes in the GTS scheme also significantly affect global simulations.

Aerosol impacts on the East Asian winter monsoon: Insights from TaiESM1 and CMIP6 simulations

“…TaiESM1 was built upon the National Center for Atmospheric Research Community Earth System Model version 1.2.2 (CESM1.2.2; Hurrell et al, 2013) and incorporated updated physical parameterizations. These enhancements encompass cloud fraction calculation using GTS (Shiu et al, 2021), the statistical‐numerical aerosol parameterization (SNAP; Chen et al, 2013), 3D radiation–topography interactions (Lee et al, 2011) and trigger functions for deep convection (Wang et al, 2015). Within TaiESM1, the aerosols can act as cloud condensation nuclei and ice nucleation particles, facilitating aerosol–cloud interactions that include cloud activation, rainout and washout.…”

A New Comprehensive Cloud Macrophysics Scheme With a Prognostic Dual‐Triangular PDF

“…Thus, some PDF-based and hybrid cloud schemes compute ice cloud fraction using a separate bulk empirical formula with an assumed supersaturation and prognoses ice cloud condensate, although liquid cloud fraction was diagnosed from the assumed PDF (S. Park et al, 2014). Similarly, Shiu et al (2021) assumed a supersaturation to diagnose ice stratus fraction, but ice water content (IWC) was still prognosed.…”

“…Similarly, Shiu et al. (2021) assumed a supersaturation to diagnose ice stratus fraction, but ice water content (IWC) was still prognosed.…”

A Neural Network‐Based Scale‐Adaptive Cloud‐Fraction Scheme for GCMs

“…Current parameterization schemes of CF can be divided into diagnostic and prognostic approaches. In the diagnostic approach, the sub‐grid CF is usually a function of the instantaneous grid‐mean properties such as atmospheric stability, relative humidity, and/or cloud water content (e.g., Shiu et al., 2021; Sundqvist et al., 1989; Xu & Randall, 1996). These schemes are easy to implement and computationally efficient but may be too simple.…”