As I noted in a post a year ago, the properties of clouds play a significant role in climate change models. Hence, it ought to be no surprise that when forecasting climate change, there’s one huge wild card: clouds. By reflecting, absorbing, and emitting radiation, clouds (and fine particles known as aerosols) play a major role in the planet’s energy balance and thus in setting Earth’s temperature. This is given the fact that Earth's climate is the product of complex interactions between the Sun, the ocean and the atmosphere. For reference, climate scientists explore these exchanges - especially between ocean and atmosphere - using general circulation models (GCM) which compute the respective circulations in ocean and atmosphere.
The resolution of the problem of narrowing uncertainty to better forecast climate change is possible only if climate scientists can better constrain the feedbacks from clouds. Enter now new research from Zelinka et al. which basically decomposes cloud feedbacks to better connect them to specific physical processes.
One primary process is radiative transfer, especially in a plane parallel atmosphere. The basic layout of the problem (especially involving clouds) is well known but difficult. Thus one would wish to investigate Rayleigh scattering in concert with standard gray atmosphere radiative transport. An equation of transfer that applies is: -dI/dt (1/k r ) = I – J
Where k is a mass scattering coefficient, r is the molecular density (e.g. in cloud cover) and J is the vector source function for a specific intensity I. If the correct Stokes parameters (I, Q, U, V) which describe degree of polarization are included, and the right incidence angle of radiation occurs, we can expect the propagation of radiant energy from the S. hemisphere to the north very effectively.
But.....one can't forget or omit diffusive reflection and re-transmission of radiation, say arising from particulates . Chandrasekhar in Radiative Transfer, (Dover Publications) shows that for angles of incidence in the range : 0.5 < i < 0.8 radian, diffusive reflection allows the radiation reflected normal to the incidence direction to actually have higher intensity than the original. (E.g. for optical depths 1.0 < < 2.0).
In effect, if conditions in the lower atmosphere incorporate such optical depths (and angles of incidence for scattering, diffusive reflection), on account of increased presence of particulates, aerosols, we will expect to find an "anomaly" say in the temperature.
For the Zelinka team, the approach entailed using so-called cloud radiative kernels along with detailed model cloud information provided by the International Satellite Cloud Climatology Project Simulator. They calculated the feedback due to 49 different types of clouds—which were divided into seven altitude categories and seven optical depth categories. (Over the past few years, kernels have become a common tool for comparing cloud feedbacks among climate models.)
In this case, the authors refined the technique to compute the amount, altitude, and optical depth feedbacks separately for upper-level and low-level clouds. Low-level clouds are those that reside in the boundary layer, the layer of Earth’s atmosphere directly influenced by its surface. Upper-level clouds reside above the boundary layer and are affected by different processes
Zelinka et al discovered that all climate models agree on the direction of three main feedbacks that accompany global warming. First, upper level clouds rise to higher altitudes in all models, warming the planet by trapping more heat. Although this was previously known, the positive feedback is actually smaller and better constrained than past research has suggested. Second, low-level cloud cover decreases in all models, warming the planet by reflecting less incoming sunlight back to space. And third, optical depth of low-level clouds increases in all models, cooling the planet by reflecting more sunlight.
The takeaway of the Zelinka et al paper is that treatment of radiative processes, transfer of radiation in clouds, is simplified by decomposing cloud feedbacks into the individual mechanisms in play. This is a crucial advance given we want to understand and constrain cloud feedbacks if we wish to more accurately forecast how our planet will react to climate change.