Evaluating and constraining models’ stratocumulus and cumulus cloud feedbacks in the tropics using satellite observations to reduce uncertainties in future climate projections

Global circulation models (GCMs) exhibit large variability in cloud
feedbacks, which limits our ability to narrow the uncertainty in models’ climate sensitivity. Part
of this uncertainty is connected to the parameterizations controlling two categories of low clouds:
stratocumulus (Sc) and shallow cumulus (Cu) clouds produced by turbulence and shallow
convective parametrizations, respectively. The two cloud types have different regional cloud
amount and feedbacks, although the mechanisms involved remain unclear. Constraining the
geographic distribution and sensitivity to environmental conditions for each of these low clouds
separately would ensure more realistic low-cloud feedbacks and help reduce the spread in model
equilibrium climate sensitivity (ECS). Such a constraint, or even a basic evaluation of the type-
resolved low clouds produced by GCMs, has been hampered a lack of both global-scale
observations that distinguish these cloud types and robust method to discriminate their regimes in
climate model.

In this project, we propose to use a new observational constraint derived from CloudSat-
Cloud-Aerosols Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) observations,
which discriminate Sc from Cu clouds, in concert with three different methods to separate Sc-
from Cu-dominated regimes, to:
• Evaluate the Sc and Cu cloud amount (geographical distributions and profiles) as simulated by
coupled model intercomparison project phase 5 (CMIP5) and phase 6 (CMIP6) climate models
as well as their associated radiative fluxes on a global scale.
• Characterize and evaluate the interannual variation of the Sc and Cu fraction (including the
vertical change) in response to surface temperature forcings and their feedbacks.
• Investigate the relationships between cloud-controlling factors and each Sc-Cu cloud type in
the observations and use it to not only evaluate how well they are replicated in CMIP5 and
CMIP6 models (informing us on how well cloud processes are represented in the PBL
parametrizations) but also to infer observationally-based future cloud feedbacks.
• Assess the impact of the Sc-Cu cloud partitioning and feedbacks on the ECS, analyze the
mechanisms behind it and estimate an observationally-constrained ECS.
• Determine what type of the parametrizations in the planetary boundary layer (turbulence and
convection) produce the most plausible Sc-Cu distribution, feedbacks and relationships with
cloud-controlling factors and for what reasons (to provide better guidance for future model
development).

Relevance of the project. We will address priority A and B of the MAPP competition 5, called
“Constraining models’ climate sensitivity”, with a type I project by narrowing the uncertainty
in the contribution of low clouds to model climate sensitivity and proposing new process-level
evaluation of model parametrizations using satellite observations. In addition, we will also address
priority C by proposing a method to observationally constrain model ECS.
The two main outcomes of the project are to reduce the uncertainty in model ECS due to low-
cloud feedback contributions and to propose new guidance for model development in specific
parametrizations (turbulence and convection). These will clearly benefit the broader community
and will directly support NOAA’s long-term research goals “weather and climate research” and “Modeling”.

Award Announcement: https://blogs.ei.columbia.edu/2020/09/01/project-clouds-climate-modeling/