Constraining and Understanding Climate Sensitivity with Process Oriented Diagnostics

The project is a holistic, bottom-up approach to understanding climate sensitivity focusing
on cloud processes. This is particularly important now because the new generation of models
participating in the Coupled Model Intercomparison Project (CMIP6) appear to show a larger
spread in estimates of climate sensitivity than expected, and in particular several prominent models
show high climate sensitivity. Among those models is the Community Earth System Model, version
2 (CESM2); this project will especially focus on understanding the high climate sensitivity in that
model. Initial work shows that there is a strong cloud response to increased carbon dioxide in
CESM2, providing a positive feedback that amplifies the warming. The proposed research delves
into the processes that drive that cloud feedback.

Using a large suite of existing simulations from the ongoing CMIP6 model intercomparison
projects, a detailed, process-oriented analysis of clouds in CESM2 will be conducted. Simulations
of the current climate will be evaluated using reanalysis and observations. A major emphasis will
be to utilize the Cloud-Feedback Model Intercomparison Project (CFMIP) simulations that include
output from the CFMIP Observational Simulator Package (COSP). This allows the simulation to be
“observed” as if by satellites, providing a consistent definition of clouds between satellite products
and model output.

Process-oriented diagnostics will be developed that focus on cloud macrophysical and micro-
physical processes. The macrophysical focus will emphasize the structure of the turbulent boundary
layer. A mixed-layer model will be used to diagnose key mixing processes at the top of the bound-
ary layer. A regime-based approach will separate cloud types, providing a fine-grained view of the
response of clouds to climate forcing. The microphysical focus will emphasize warm rain processes
and aerosol-cloud interactions, and how each change with climate forcing. The same regime-based
approach will be applied. The macrophysical and microphysical factors may work in concert or
oppose each other in different regimes and under different forcing, and the diagnostics developed
here will expose these balances.

The process-oriented diagnostics will be developed for CMIP-style model output. That will
allow them to be applied to the multi-model ensemble to better understand cloud processes and
the relationship to climate sensitivity. The comparisons with observations provide physically-based
constraints on the model processes; such constraints may be used make a more confident statement
about the range of plausible climate sensitivity in current models, and will also directly inform
model development activities. The diagnostics will be developed using the Model Diagnostics Task
Force (MDTF) framework, and they will be submitted to the Task Force for consideration to be
included in the package. Therefore this project directly contributes to NOAA’s long-term climate
research goals by elucidating the role of clouds in a changing climate, providing observation-based
constraints on the physical processes that contribute to uncertainty in climate projections, and
providing new information and tools that can be used by the broad scientific community and in
assessments of climate change such as the National Climate Assessment.