In the Met Office's global weather prediction it is essential to have continuous measurements of the vertical structure of temperature and humidity in the atmosphere. This is routinely carried out by satellite instruments operating at infrared wavelengths, and our capabilities in this area will improve with the launch of the European Infrared Atmospheric Sounding Interferometer (IASI) in 2006. IASI will sound the atmosphere over thousands of wavelength channels.

In preparation for the launch, the Atmospheric Radiation Research Group has been involved in testing our ability to model the infrared radiance spectra IASI will produce. Only by understanding the emission and absorption characteristics of the atmosphere can we accurately infer temperature and humidity profiles.

The Airborne Research Interferometer Evaluation System (ARIES) on the FAAM BAe 146-301 aircraft is similar in design to IASI, and allows us to measure infrared radiance spectra at the same time as gathering in situ data on the state of the atmosphere. In this way we can critically assess our ability to retrieve temperature and humidity information using infrared spectra.

One important uncertainty in infrared radiance modelling is the strength of the water vapour continuum, an underlying feature that can be considered additional to the usual emission lines of water in the atmosphere. Uniquely with ARIES we can fly at low altitudes viewing up through the atmospheric column to measure the strength of the continuum at around 800 to 1,200 cm^-1 (8.3 to 12.5 mm wavelength). The figures below summarise our results.

Fig 1a: our results predict a weaker continuum than the standard continuum model (CKD 2.4). This conclusion is supported by laboratory results shown as the data points with error bars.

Fig. 1b: ARIES brightness temperature spectrum and residual differences between measurements and simulations; we obtain much better agreement with the measurements by using our derived reduced-strength continuum.

Recently the CKD 2.4 water vapour continuum model has been superseded by the MT_CKD model, which incorporates some reduction in intensity in the atmospheric window as we have demonstrated.

Another important quantity involved in infrared sounding from space is the emissivity of the surface. Essentially the emissivity dictates how efficient the surface is at radiating heat, and needs to be known if we are to measure surface temperatures from space.

Sea-surface temperatures (SSTs) are routinely measured using wavelength channels where the atmosphere is relatively transparent. Until recently the emissivity was assumed to be constant for all water temperatures, but we observed differences in the sea-surface emission when flying over the tropical ocean and the Baltic Sea. In order to investigate this further we performed experiments on the ground using ARIES.

• We used a tank of calm, flat water to remove the effect of surface roughness (waves caused by wind) that complicate such measurements in the real ocean.
• We carefully controlled the temperature and salinity of the water to monitor their effects.
• We measured both the water-emitted radiance and the downwelling sky radiance to remove the reflected part of the radiation and so measure the emissivity component accurately.

Our results for pure deionised water are shown below.

Fig. 2a: the water surface emissivity spectrum shows a consistent peak at around 900 cm-1 in the atmospheric window region, but as the water temperature is reduced differences become apparent, particularly around 800 cm-1 as shown in the lower (obs-calc) difference plot.

Fig. 2b: if the standard model emissivity, rather than our measured emissivity, is used to derive an SST from infrared radiance measurements then systematic biases in SST result. These biases are largest at the coldest sea temperatures so that at near-freezing temperatures an error of 0.4 °C is introduced if frequency channels around 800 cm-1 are used in the SST retrieval.

The effect of dissolved salts was found to be less significant than temperature. Using a combination of results obtained with ARIES over ocean (on the aircraft) and during the ground-based experiments we were able to derive a temperature-dependent  refractive index for salt water which we anticipate will be reliable over the range of temperatures encountered over the world's oceans. We expect that satellite measurements of SST will be more accurate as a result.

References

Taylor, J.P., Newman, S.M., Hewison, T.J. and McGrath, A., 2003: Water vapour line and continuum absorption in the thermal infrared — reconciling models and observations, QJR Meteorol Soc, 129, 2,949-2,969.

Newman, S.M., Smith, J.A., Glew, M.D., Rogers, S.M. and Taylor, J.P., (2005): Temperature and salinity dependence of sea surface emissivity in the thermal infrared, Submitted to QJR Meteorol Soc.