I notice that some posters have been discussing water vapor feedbacks and thought that I’d chip in a little. One aspect of water vapor’s role in the climate system that I find intriguing is absorption of near infrared (NIR) and visible solar radiation by water vapor. This is a topic of very active research by molecular spectroscopists, for several reasons, not least being the very significant discrepancy between observed atmospheric absorption of solar radiation and modeled absorption in the IPCC TAR models (up to 30 wm-2 as compared to 3 wm-2 for CO2 doubling).
Part of the recent interest was prompted by the discovery of a delicious clerical error under-transcribing the NIR absorptions in HITRAN-96 by an amount greater than the impact of 2xCO2. I’ll locate some references to this. New experimental work has shown that NIR absorption by water vapor was under-estimated by much more than even these clerical errors. Jonathan Tennyson, Professor of Physics, Head of Department, Department of Physics and Astronomy, University College, London and an extremely eminent near infrared spectroscopist, provided an interesting popular summary of some of these issues in 2003.
A.N. Maurellis and J. Tennyson, The climatic effects of water vapour, Physics World, 29-33, April 2003) here said:
There are a number of popular misconceptions about the greenhouse effect, notably that it is a bad thing. On the contrary, the greenhouse effect is a significant factor in making the Earth habitable. Without it the average temperature on Earth would be lowered by about 30 K, which would make most of the planet’s surface decidedly chilly. Furthermore, it is the water vapour in the lower 10 km or so of the atmosphere, rather than man-made carbon-dioxide emissions, that contributes most to this warming effect.
But when the absorption values in the HITRAN database are used in model-atmosphere calculations, the results are disturbing. For clear skies, the models predict that the atmosphere absorbs much less sunlight than is measured by a variety of satellite and aircraft. The difference between the predictions and the measurements can be as large as 30 W m-2. (see "Radiation budget is called to account" by A Maurellis Physics World November 2001 pp22-23). This problem has become known as the absorption anomaly. And there are even worse problems in understanding absorption models when the sky is cloudy. Not all models underestimate the amount of atmospheric absorption because some physicists choose to add extra absorption to their models to mop up the surplus radiation. However, the physical cause of the missing clear-sky absorption and its exact wavelength distribution remain unresolved, and a source of fertile speculation. Everyone’s favourite molecule is always a candidate. The precise effect of these absorption bands is hard to determine, despite the best efforts of many talented and dedicated scientists…
Experiments that were performed by Roland Schermaul and the late Richard Learner at Imperial College in London in 2001 have cast previous measurements of the absorption spectrum of water into considerable doubt. The study was motivated by the European Space Agency (ESA), which was concerned that the uncertainty in water-vapour data was preventing important information on trace molecules in the atmosphere from being obtained. Schermaul and co-workers used the Molecular Spectroscopy Facility at the Rutherford Appleton Laboratory in the UK to study the absorption of light by water vapour in air at wavelengths that varied from the near infrared to the orange. They found that the strong spectral lines absorbed significantly more light – between 5% and 25% – than previous laboratory measurements had suggested. This conclusion was given partial support by first-principle quantum-mechanical calculations, which can be used to estimate the strength of these absorptions….
These measurements were put into atmospheric models by Joanna Haigh’s group at Imperial College to find out if they could explain the absorption anomaly. When the strength of the strong water absorption lines was increased in the model, the absorption of incoming sunlight rose by about 8 W m-2. This increased by a further 3 W m-2 when the weak line parameters that were measured by Schermaul and co-workers were included. Together these increases represent about half of the absorption anomaly. Unfortunately, however, the situation is not quite this straightforward.
There is a recent article (April 2005) by Martin Wild in GRL considering the handling of NIR absorption in recent GCMs. I’ll post up some info on this and some other NIR references.