| 英文摘要: | Constraining climate sensitivity is a top priority for climate science. Now research shows that the details of how stratospheric ozone is represented in models can have a strong influence on warming projections. To predict how climate will change we need to know just two things: the future forcings on the Earth system, and the feedbacks that will ensue. The first requires forecasts of socioeconomics, technology, and human behaviour to be translated into changes in the planetary energy budget; for example, via changes in anthropogenic emissions and atmospheric composition. The second requires understanding of how processes interact to either amplify or damp climate response to the forcing, together with the timescales over which they operate. This response is encapsulated by a parameter called the climate sensitivity. Warmer air can (and generally does) accommodate more water vapour molecules. As water vapour is a greenhouse gas, any warming is amplified: this represents a positive feedback — it increases climate sensitivity. Several other feedbacks operate in the Earth system, including those associated with sea ice and clouds; most are positive. Negative feedbacks appear to be less common. Writing in Nature Climate Change, Peer Nowack and co-authors2 report a large negative feedback, which causes a significant (in the order of 20%) reduction in climate sensitivity, resulting from a more comprehensive model representation of stratospheric ozone chemistry. Quantifying climate sensitivity could hardly be more important. Equilibrium climate sensitivity (ECS) is defined by the surface warming in a world with doubled pre-industrial atmospheric CO2 levels, once the system has returned to equilibrium. Current estimates of ECS are wide1: 1.5–4.5 °C. If Earth's climate sensitivity is high, then rapid and severe cuts in emissions are needed to have any chance of avoiding dangerous climate change3. Climate sensitivity can be estimated in a variety of ways, from both observations and models1. One commonly employed method involves subjecting a range of climate models to an abrupt quadrupling of atmospheric CO2, and following their simulated surface temperature response over at least the next century4. Different representations of key Earth system processes, such as convection, cause each model to have an individual response to the atmospheric forcing. These result in different simulation of feedbacks, and consequently some models project more warming than others. Together with poorly constrained future emissions, this range of model responses generates uncertainty in climate projections. Ever increasing computing power offers climate modellers the opportunity to make their models more accurate representations of the real world. Development has followed two (overlapping) paths: increasing spatial resolution; and increasing the number and/or complexity of processes represented. The latter path has resulted in the development of 'Earth system models' — these incorporate chemical and biological processes into the (almost entirely) physical framework of the earlier generation of climate models. Broadening the scope of Earth system models allows a wider range of potential feedbacks to spontaneously emerge — including climate feedbacks beyond those of a purely physical nature, first expounded by James Lovelock in his pioneering Gaia hypothesis5. Nowack et al.2 use an Earth system model to investigate how inclusion of interactive chemistry influences climate sensitivity. Interactive chemistry allows distributions of radiatively active gases (in this case, ozone) to change in step with the climate, rather than prescribing them. By conducting the same experiment in two model versions (one with and one without interactive stratospheric chemistry), they isolate the impact of interactive stratospheric chemistry, and find that it induces a strong negative climate feedback. The mechanism is linked to the Brewer–Dobson circulation6, 7, which moves stratospheric air from the tropics to higher latitudes (Fig. 1). Models consistently predict a strengthening of this circulation as surface climate warms6. A stronger circulation lifts the tropical tropopause (the boundary between the troposphere and stratosphere), decreasing ozone in the tropical lower stratosphere. As ozone is a greenhouse gas, less ozone leads to surface cooling — a negative feedback. Temperatures also decrease locally in the layer of the atmosphere where the ozone is reduced, around the tropopause. Colder tropopause temperatures enhance the freeze-drying of air entering the stratosphere, reducing stratospheric water vapour, generating yet more surface cooling — a further negative feedback. However, decreased upper tropospheric temperatures promote cirrus cloud formation. More high cloud leads to surface warming, so the increase in cirrus represents a positive feedback. The overall impact of all these processes is dominated by the reduction in ozone and water vapour, resulting in a net negative feedback, with Nowack et al.2 finding about 20% less warming in their experiment with fully interactive chemistry.
|
