Investigating Cloud Feedbacks in Earth System Models

The coupling between clouds, radiation and the circulation represents a key source of uncertainty in model predictions of climate change.

climate change

A change in the average conditions — such as temperature and rainfall — in a region over a long period of time.

Source: NASA Climate Kids Source

Interactions between clouds and radiation play a fundamental role in regulating the large-scale atmospheric circulation which, in turn, governs the distribution of clouds and precipitation. Climate models are very sensitive to how such clouds are coupled to the larger-scale circulation. Cloud-circulation feedbacks represent a major source of uncertainty in our understanding of climate sensitivity, particularly for tropical low clouds, which are strongly modulated by changes in the large-scale circulation.

Convective aggregation refers to the tendency for tropical convection

convection

The rising of warm air and the sinking of cool air. Convection can cause local breezes, winds, and thunderstorms.

Source: NASA earth observatory Source

to self-organize into limited regions. It has been primarily studied in numerical simulations of radiative convective equilibrium under idealized conditions (e.g., uniform surface temperature). Models generally predict increased convective organization in a warming climate; however, there remains widespread disagreement about the nature of this dependence, reflecting the uncertainty in representing interactions between clouds, radiation, moisture, and the circulation. Moreover, few studies have examined its relative importance when considered in the presence of realistic, rather than idealized, boundary conditions (e.g., SST gradients, wind shear, etc.).

Recent observational studies suggest that changes in convective aggregation modulate both the radiation budget of the tropics and the intensity of extreme precipitation events. Observations indicate increased aggregation is associated with a reduction of high cloud cover, drying of the free troposphere, an increase in low cloud amount and an increase in extreme precipitation. Such changes are critical to simulating climate change in the tropics, yet the fidelity of coupled ocean-atmosphere and Earth System Models

earth system model (ESM)

A model that simulates how chemistry, biology, and physical forces work together. These models are similar to, but much more comprehensive than, global climate models.

Source: U.S. Department of Energy Source

(ESMs) in reproducing these changes remains unclear.

Here we seek to investigate how the coupling of clouds, radiation and dynamics modulates both the climate sensitivity as well as the response of extreme weather events to changes in climate. In particular, we seek to answer the following questions:

   i) Are radiative feedbacks to convective aggregation important under realistic boundary conditions?

   ii) How do cloud-circulation feedbacks influence the climate sensitivity of models?

   iii) How do cloud-circulation feedbacks influence the distribution and response of weather extremes?

   iv) How can observations be used to evaluate and constrain the representation of these feedbacks?

To accomplish these tasks, we will first compare ESM simulations from CMIP6 against observations to evaluate their ability to reproduce observable co-variations in the relevant feedback quantities. This analysis will then be complimented by a series of sensitivity experiments in which the vertical distribution of radiative heating is manipulated to suppress synoptic-scale cloud-circulation feedbacks using the atmospheric models from the DOE and GFDL ESM frameworks. Initial “proof of concept” experiments highlight the potential utility of this framework for isolating cloud-radiative interactions. These experiments will help to quantify the feedbacks and establish causality of the relationships diagnosed from the CMIP6 simulations. The last step of this work will utilize global-scale observations across a range of time scales and Machine Learning (ML)

machine learning (ML)

Machine Learning (ML) refers to the field and practice of using algorithms that are able to “learn” by extracting patterns from a large body of data.

Source: Center of Excellence Source

techniques to identify emergent constraints on the representation of these feedbacks in models.