Shifts in Atmospheric Circulation Alter Global Clouds and Affect Climate Sensitivity

The interaction between atmospheric circulation in the tropics and sub-tropics and cloud structure is highly correlated.

Interestingly, changes in the circulation and the consequent shifts of cloud cover lead to differing warming/cooling effects in the northern and southern hemispheres. How well climate models simulate the southern hemisphere effects has implications for the models’ climate sensitivity.

The image above shows a particular day’s snapshot of the Earth from the NASA Terra satellite. Examining the structure of the global cloud field one can see, a long line of bright, dense clouds near the equator in a region known as the Intertropical Convergence Zone (ITCZ), more clearly evident over the tropical Atlantic and East Pacific oceans. Moving away from the tropics in both hemispheres, one encounters the subtropical zone of relatively low cloud coverage, occupied by decks of broken clouds over ocean and clear desert areas over land. Finally, in the middle and high latitudes of both hemispheres one can see the comma-shaped storm clouds familiar from the everyday weather reports. This structure of the global cloud field is a creation of the Earth’s atmospheric circulation.

Atmospheric circulation, when examined using a simplified, two-dimensional view (such as in Fig. 2, below), is dominated by two major features. The first is a large feature called the Hadley cell, which lifts air in the ITCZ, moves it at high altitudes towards the poles, and sinks it again to the surface in the subtropical regions. The second feature is a very strong river of air, known as the jet stream, that flows from west to east in the middle latitudes of each hemisphere. The meanders of the jet stream produce the storm tracks that are the major weather makers in the midlatitude regions.

The second figure shows a zonal average of this circulation and the locations of these features. The subsiding zones at latitudes between 20° and 30° north and south are noted by the letter ‘H’, the jet stream is in each hemisphere is marked with a dot, and the storm tracks are noted with an ‘L’. In this figure, the circulation is superimposed on the distribution of the world’s clouds, derived from NASA Cloudsat satellite observations. It is apparent how they relate to the circulation features. The narrow zone of uplift in the tropics produces high, thick clouds in the ITCZ. The areas of subsidence in the subtropics produce extensive fields of low clouds, more extensive and deep in the southern than in the northern hemisphere, while the storms embedded in the jet stream produce deep, high clouds that extend throughout the Earth’s troposphere.

Fig. 2: Circulation and clouds in the climate. The color background denotes cloud cover. Location of the subsiding branches of the Hadley Cell are denoted by an ‘H’. The jet streams and storm tracks are marked by the dots and the letter ‘L’ respectively.

Observations of the past 35 years indicate that, as the Earth has warmed, these circulation features are moving towards the poles. The Hadley cell shows a clear signal of poleward expansion, while poleward movement is present but less clear in the jet stream and mid-latitude storm tracks. This is broadly consistent with climate model simulations of future climate warming, which predict significant poleward movement of all the major circulation features.

The amount of planetary warming for a doubling of CO2 is called the climate sensitivity, which scientists have estimated is between 2° and 4.5°C. We know that much of the uncertainty in this number is related to how clouds will change with warming. Increases in high, thin clouds would suggest a larger sensitivity, while increases in low, thick clouds would make the value smaller. A critical question, then, is how the cloud field will respond to the circulation shifts we are already seeing.

In a paper published last year team of scientists from NASA/GISS, Columbia University, and the University of Virginia, correlated observations of cloud properties and circulation indices over the past 35 years. We found that the two quantities that correlate significantly and consistently in all ocean basins and seasons are the Hadley cell extent and the high cloud field: when the Hadley cell edge moves poleward the high cloud field also shifts towards the poles and vice versa. However, this coordinated movement does not have the same effect in the two hemispheres. In the northern hemisphere, the poleward movement of the high clouds opens up a “cloud curtain” that lets more sunlight into the ocean surface, thus producing warming at the surface. But in the southern hemisphere, the poleward contraction of the high clouds is balanced by an expansion of the already extensive low cloud decks, which ends up blocking more sunlight and producing a small surface cooling.

In a follow-up paper, we examined how climate models simulate the interactions between clouds and the atmospheric circulation, and what the implications might be for the models’ climate sensitivity. We found that, for the current climate, most models tend to have climatological Hadley cells that are too narrow and, therefore, underestimate the extent of the wide deck of low clouds associated with the southern subsiding branch. This means that when the Hadley cell extends poleward as the climate warms, the contraction of the high clouds remains largely unbalanced and the surface of the southern hemispherer oceans warms. Since all models significantly expand their Hadley cells in climate warming simulations, this cloud-circulation interaction leaves a significant imprint on the radiative impact of the clouds.

Fig. 3: Short-wave cloud radiative effect. The effect is shown for 4×CO2 for the CMIP5 models as a function of latitude in the southern summer season. Differences in models are dominated by the southern hemisphere changes. (Image Credit: NASA/GISS/Lipat et al.)

The third figure shows the model-simulated change in solar cloud radiative effect with climate warming, with positive changes implying that clouds have an amplifying effect. It can be seen that most models produce radiative warming in the southern mid-latitudes as they expand their Hadley cells poleward. The differences in how strongly the modes simulate this effect is connected to climate sensitivity. Models with more expansive climatological Hadley cells tend to warm this region less or not at all, and tend to have relatively lower climate sensitivities. Models with excessively narrow climatological Hadley cells tend to warm this region more and tend to have higher climate sensitivities. Overall, models that better represent this cloud-circulation interaction have a climate sensitivity near 3°C compared — models that do not have climate sensitivities between 4° and 5°C.

It is important to remember that climate sensitivity is determined by the accumulated effects of competing climate feedbacks. For example, in addition to showing the future climate solar warming zone, Fig. 3 shows a zone of strong solar cloud radiative cooling, located between 50° and 80°S. This cooling is caused by the fact that that, as climate warms, clouds made of large ice particles are replaced by clouds made of more numerous and smaller water particles that reflect more solar radiation. A recent study used NASA satellite observations to test the skill of climate models in simulating this cloud-type transition, and found that high sensitivity models simulate it more accurately, while low sensitivity models tend to overemphasize its climate cooling effect. The challenge, therefore, for climate scientists is to try and synthesize the evaluation of climate models with diverse skills in simulating important climate processes, using observations of those processes to assess their relative importance in determining climate sensitivity.

Further Reading
NASA Feature, 5/2016: Expanding Tropics Pushing High Altitude Clouds Towards Poles

References
Lipat, B.R., G. Tselioudis, K.M. Grise, and L.M. Polvani, 2017: CMIP5 models’ shortwave cloud radiative response and climate sensitivity linked to the climatological Hadley cell extent. Geophys. Res. Lett., early on-line, doi:10.1002/2017GL073151.

Tselioudis, G., B. Lipat, D. Konsta, K. Grise, and L. Polvani, 2016: Midlatitude cloud shifts, their primary link to the Hadley cell, and their diverse radiative effects. Geophys. Res. Lett., 43, no. 9, 4594-4601, doi:10.1002/2016GL068242.

Please address all inquiries about this research to Dr. George Tselioudis.