Climate Change 2001:
Working Group I: The Scientific Basis
Other reports in this collection

7.2 Atmospheric Processes and Feedbacks

7.2.1 Physics of the Water Vapour and Cloud Feedbacks

Any increase in the amount of a greenhouse gas contained in the Earth's atmosphere would reduce the emission of outgoing long-wave radiation (OLR) if the temperature of the atmosphere and surface were held fixed. The climate achieves a new equilibrium by warming until the OLR increases enough to balance the incoming solar radiation. Addition of greenhouse gases affects the OLR primarily because the tropospheric temperature decreases with height. With a fixed temperature profile, increasing the greenhouse gas content makes the higher parts of the atmosphere more opaque to infrared radiation upwelling from below, replacing this radiation with OLR emitted from the colder regions.

Determination of the new equilibrium is complicated by the fact that water vapour is itself a potent greenhouse gas, and the amount and distribution of water vapour will generally change as the climate changes. The atmospheric water vapour content responds to changes in temperature, microphysical processes and the atmospheric circulation. An overarching consideration is that the maximum amount of water vapour air can hold increases rapidly with temperature, in accord with the Clausius-Clapeyron relation. This affects all aspects of the hydrological cycle. Unlike CO2, water vapour concentration varies substantially in both the vertical and horizontal. An increase in water vapour reduces the OLR only if it occurs at an altitude where the temperature is lower than the ground temperature, and the impact grows sharply as the temperature difference increases. If water vapour at such places increases as the climate warms, then the additional reduction in OLR requires the new equilibrium to be warmer than it would have been if water vapour content had remained fixed. This is referred to as a positive water vapour feedback.

Clouds are intimately connected to the water vapour pattern, as clouds occur in connection with high relative humidity, and cloud processes in turn affect the moisture distribution. Clouds affect OLR in the same way as a greenhouse gas, but their net effect on the radiation budget is complicated by the fact that clouds also reflect incoming solar radiation. As clouds form the condensation releases latent heat, which is a central influence in many atmospheric circulations.

The boundary layer is the turbulent, well-mixed shallow layer near the ground, which can be regarded as being directly moistened by evaporation from the surface. In the boundary layer, the increase in water vapour with temperature in proportion with the Clausius-Clapeyron relation is uncontroversial. Observations (e.g., Wentz and Schabel, 2000) clearly show a very strong relation of total column water vapour (precipitable water) with surface and tropospheric temperature. Because the boundary-layer temperature is similar to that of the ground, however, boundarylayer water vapour is not of direct significance to the water vapour feedback. Furthermore, half of the atmospheric water vapour is below 850 mb, so measurements of total column water have limited utility in understanding water vapour feedback. The part of the troposphere above the boundary layer is referred to as the "free troposphere". Water vapour is brought to the free troposphere by a variety of mixing and transport processes, and water vapour feedback is determined by the aggregate effects of changes in the transport and in the rate at which water is removed by precipitation occurring when air parcels are cooled, usually by rising motions.

The complexity of water vapour radiative impact is reflected in the intricate and strongly inhomogeneous patterns of the day-to-day water vapour distribution (Figure 7.1a). The very dry and very moist regions reveal a strong influence of the large-scale dynamical transport. Model simulations exhibit similar patterns (Figure 7.1b), with a notable qualitative improvement at higher resolution (Figure 7.1c). Understanding the dominant transport processes that set up those patterns, and how they can be affected by a modified climate, should help assess the representation of the water vapour feedbacks in the corresponding region found in model simulations.



Figure 7.1: Comparison between an observational estimate from satellite radiances and two model simulations of the complex structure of midtropospheric water vapour distribution for the date May 5. At any instant, water vapour is unevenly distributed in the atmosphere with very dry areas adjacent to very moist areas. Any modification in the statistics of those areas participates in the atmospheric feedback. The observed small-scale structure of the strong and variable gradients (a) is not well resolved in a simulation with a climate model of the spatial resolution currently used for climate projections (b), but simulated with much better fidelity in models with significantly higher resolution (c). (a): Distribution of mean relative humidity in layer 250 to 600 mb on May 5, 1998, as retrieved from observations on SSM/T-2 satellite (Spencer and Braswell, 1997). Missing data are indicated by black areas and the retrieval is most reliable in the latitude band 30°S to 30°N. (b): Relative humidity at about 400 mb for May 5 of an arbitrary year from a simulation with the GFDL R30L14 atmospheric general circulation model used in the AMIP I simulation (Gates et al., 1999; Lau and Nath, 1999). In small polar areas (about 5% of the globe) some relative humidities are negative (set to zero) due to numerical spectral effects. (c): Relative humidity at 400 mb from the ECHAM4 T106 simulation for May 5 of an arbitrary year (Roeckner et al., 1996; Wild et al., 1998).

Continues on next page



Other reports in this collection