GFDL - Geophysical Fluid Dynamics Laboratory

GFDL’s Cloud-Climate Initiative

Lead: Ming Zhao

Members: Leo Donner, Huan Guo, Isaac Held, S-J Lin, Yi Ming, Max Popp, V. Ramaswamy, Levi Silvers, Michael Winton, Baoqiang Xiang, Rong Zhang

Background

Clouds cover approximately two-thirds of the Earth’s surface and are often organized into coherent systems by large-scale atmospheric flows. Clouds can warm the Earth by trapping outgoing longwave infrared radiative flux at the top of the atmosphere. Clouds can also reflect shortwave solar radiative flux back to space and cool the Earth. The net effect of the two competing processes depends on the height, type, and the optical properties of the clouds.

Given the large magnitude of the warming and cooling effects of clouds, they have potential to cause significant climate feedback. The sign of the feedback on climate change depends on how sensitive the properties are that govern both the longwave and shortwave radiative fluxes. Thus, estimates of cloud feedback require process-level understanding and modeling of the nontrivial factors on which clouds depend. Moreover, since clouds modify the general circulation and hydrologic cycle through their interactions with the atmosphere, ocean, and land, comprehensive global climate models are seen as a crucial tool in our quest for an adequate understanding of the interactions between clouds and climate.

The figure below shows that clouds occur most frequently over the tropical precipitation belts and the middle-latitude oceanic storm track regions, with the continental desert and central subtropical oceans being relatively cloud-free. Clouds are composed of liquid at temperatures above zero, ice below about -38C, and either or both phases at intermediate temperatures. Since clouds extend higher in the tropics, ice and mixed-phase clouds are important throughout all latitudes. At any given time, most clouds are not precipitating, but precipitation can be crucial in regulating clouds and associated moist turbulence processes.

From the IPCC AR5 report (Chapter 7, Fig.7.5): (a) annual mean cloud fractional occurrence. (b) annual zonal mean liquid and ice water path. (c-d) latitude-height distributions of annual zonal mean cloud and precipitation occurrence (magnitude of precipitation occurrence is doubled to make use a common color scale).

Goals

GFDL created a Cloud-Climate Initiative in 2012, to improve our understanding of the role of clouds in the Earth’s present climate, and its future changes at both global and regional scales. In particular, this focused effort is designed to explore the connections and feedbacks between convection, clouds, radiation, circulation, the hydrological cycle, and climate sensitivity, using a hierarchy of models in combination with observational and reanalysis data.

One of our goals is to develop strategies for global and process-level evaluation of convection and cloud parameterizations in climate models. GFDL scientists hope to develop physically-based convection and cloud parameterizations, to increase physical realism in the interactions among convection, clouds, and climate and thus systematically attempt to narrow climate model estimates of climate sensitivity.

Current Activities

GFDL scientists are exploring the causes for the uncertainty in model-simulated cloud feedback and climate sensitivity, in order to gain a process-level understanding of clouds and cloud feedback sensitivity in our climate system. Exploring model simulations with realistic initial and boundary conditions, and using observations to evaluate modeled clouds, will give us insights into the causes of cloud biases in climate models. Evaluating model-simulated cloud variability against observations, at a wide range of temporal and spatial scales, will help us identify emergent constraints on modeling clouds and cloud feedback.

GFDL scientists use idealized (i.e., simplified) frameworks to explore convection-cloud-circulation-climate interactions. One example is an aqua-planet model, with the land entirely removed. This allows scientists to isolate processes that are fundamental to modeling clouds and climate, then build on that understanding with added complexity. We have developed a hierarchy of models, from simple to complex, to bridge the gap between models and theories.

Finally, GFDL scientists participate in relevant national and international cloud-climate projects to strengthen connections between GFDL and worldwide cloud-climate research communities. Using community-based model inter-comparison experiments, such as Cloud Resolving Model / Large Eddy Simulation, informs our own experiments and improves our understanding of cloud feedbacks.

Challenges of Modeling Cloud-Climate Interactions

Cloud processes span a large range of scales from sub-micrometer to thousands of kilometers. This range of scales cannot be explicitly resolved with numerical models for the foreseeable future. Cloud processes also involve many complex interactions among cloud microphysics, precipitation, moist turbulence, latent heat release, circulation, and radiation. The representation of moist turbulence and cloud microphysics are especially challenging in climate models, without detailed knowledge of the sub-grid scale turbulence statistics.

There are many cloud regimes coexisting around the globe, reflecting diverse meteorological environments (e.g., shallow/deep, liquid/ice, ascent/descent, ocean/land, low/middle/high latitude or altitude, day/night, flat/mountainous, etc.). The processes and their interplay often differ significantly between regimes, making a unified representation especially difficult. The wide variety of cloud regimes are all likely important to the overall Earth radiation budget and need to be accounted for accurately.

These challenges also apply to theories and observations. Thus, a comprehensive modelling and observational strategy in combination with theories is necessary to advance our understanding.

How Clouds Might Respond to Global Warming

Recent studies suggest several cloud feedback mechanisms, which appear to be consistent among different climate models, and are supported by other lines of study. As the climate warms, high clouds are expected to rise in altitude and thereby exert a stronger greenhouse effect. In a warmer climate, middle and high-level cloud cover tend to decrease. Storm tracks shift poleward, drying the subtropics and moistening the high latitudes, which tends to cause positive feedback by shifting cloud cover to latitudes that receive less sunshine. Most climate models predict decreases in low cloud amount, especially in the subtropics. Over middle and high latitudes, climate models suggest that warming-induced transitions from ice to water clouds may result in more opaque clouds.

Despite these robust cloud responses, large uncertainties remain. In particular, it is not clear to what extent the reduction of high cloud cover may compensate for rising high cloud altitude. The response of cloud water, its phase transition and radiative effects, are sensitive to representations of cloud precipitation microphysics, which remains highly uncertain and difficult to constrain in climate models with observations. Further, the mechanisms causing reductions in low cloud amount in models may differ significantly among models, and lack a well-accepted theoretical basis. Cloud process models suggest a variety of potentially opposing response mechanisms that may account for the current differences among climate models. More research is needed to narrow down cloud feedback uncertainties, as well as possible emergent constraints.