CLIMATE Feedback (CF) is a direct or indirect partial response (sensitivity) of any component of the climate system (for example, clouds, atmospheric chemical composition, or atmospheric water vapor) back on the Earth's near-surface temperature (NST) that was initially offset by an external to the climate system forcing. An initial disturbance of the NST comes from a change in the physical properties of the surface (for example, due to human agricultural activities) or atmosphere (for example, an industrial increase of greenhouse gases), or occurs due to the presence of an external to the system's heat source (such as urban heat islands). When the NST is initially forced, a feedback either amplifies (positive feedback) or damps/controls/scales (negative feedback) the effect of the disturbance back on the NST.
CF should be differentiated from radiative feedbacks (RF), which represent a response of the NST back to the radiative balance at the top of the atmosphere (TOA), which was initially offset by an external forcing. The radiative balance (RB) at the TOA is the difference between the incoming into the Earth's solar radiation, and the outgoing terrestrial radiation. Both CF and RF are components of a climate feedback loop's network. The system's net feedback loops effect (netFLE) accounts for all possible influences of the climate variables on each other and is compounded with direct and indirect feedback loops.
There are many CFs that are recognized in the Earth's climate system of interactions. Among them are water vapor feedback, lapse-rate feedback, ice/ snow albedo feedback, carbon cycle feedback, bio-geochemical cycles feedback, dynamical feedback, ocean circulation and marine life feedbacks, and cloud feedbacks. In most cases, feedbacks can be expanded further to a set of other feedbacks in accordance with the physical properties of a climate variable in relation to the NST. Thus, a cloud feedback is expandable to changes in cloud cover, cloud top temperature and/ or height, cloud optical thickness, or cloud droplet size. It is recognized that, for the climate system, the most important CF is water vapor feedback, which is closely coupled to lapse-rate and cloud feedbacks.
There are direct (one-to-one relationships between temperature and a climate variable) and induced (via other variables) CFs. Another term for induced feed back is indirect feedback. Due to the atmospheric, oceanic, and land processes interactions, most of which depend on temperature, the direct CFs compete with, and are enriched by, the induced ones. Sometimes, it is possible to find a scale in time (such as annual or decadal) and space (such as regional or global) where a separation of the climate variables into fast and slow varying ones is possible, and the climate system can be simplified to understand and predict its evolutionary paths. Slow varying variables define slow feedbacks. Fast feedbacks, acting along a chain of interactions, cancel each other, affecting the tendencies of the climate system evolution. An example of a fast negative CF is a lapse-rate feedback (sometimes termed atmospheric temperature profile feedback), which establishes the atmospheric temperature profile and the magnitude of the near-surface temperature in response to surface heating in a matter of hours. An example of a slow positive CF is a global ice-albedo feedback, which is active over many centuries and brought the Earth to the ice ages and back.
Strength of the feedbacks is defined by climate system composition (for example, how many elements of the climate system are under consideration), an initial state (for example, the recent climate, or an ice age), and the nature of the relationship among climate variables spanned over spatial and time domains. For example, an atmospheric feedback initiated by surface warming due to an increase in carbon dioxide is different in dry and moist atmospheres: the presence of water vapor in the atmosphere increases the strength of the dry atmosphere positive feedback on the surface temperature. Feedbacks initiated in a cloudy and/or polluted atmosphere are different from ones initiated in a clear (no clouds) and clean (no aerosols) atmosphere. Feedbacks resulting from a volcano confined to the polar vortex region (or just to one hemisphere) will be different from the feedbacks resulting from a tropical volcano eruption.
A general measure of CFs is the climate sensitivity (CS), or its reciprocal, a climate feedback parameter (CFP). The CS depends on a netFLE and where climate forcing is in place. The CS and CFP are easily estimated from the numerical models as they represent a closed description of the model climate state's evolution and they allow for executing "sensitivity" experiments with a model. Differences in the model
CS estimations are explained by implementation of a slightly different set of feedbacks (such as that related to clouds) that amplify or dampen the initial perturbation. It is difficult to infer CS from the observations, as they do not supply a complete and continuous mirror of the real world, and each observational climate state carries footprints of numerous non-attributed forcings from the present and past. Without a model, it is difficult to recognize a single feedback or a single feedback loop, or to validate an estimated model feedback with observations. However, observations are the only source to validate model parameterizations, which are approximations of the climate system feedbacks.
SEE ALSO: Biogeochemical Feedbacks; Climate Forcing; Climate Models; Dynamical Feedbacks; Evaporation Feedbacks; Ice Albedo Feedback.
BIBLIOGRApHY. D.A. Lashof, et al., "Terrestrial Ecosystem Feedbacks to Global Climate Change," Annual Reviews of Energy and the Environment (v.22, 1997); Panel on Climate Change Feedbacks, Climate Research Committee and National Research Council, "Understanding Climate Change Feedbacks" (National Academy of Sciences, 2003); Margaret Torn and John Harte, "Missing Feedbacks, Asymmetric Uncertainties, and The Underestimation of Future Warming," Geophysical Research Letters (v.33/10, 2006).
Natalia Andronova University of Michigan
Was this article helpful?