Atmospheric observations of the water vapour continuum in the near-infrared windows between 2500 and 6600 cm−1

Abstract

Abstract. Water vapour continuum absorption is potentially important for both closure of the Earth's energy budget and remote sensing applications. Currently, there are significant uncertainties in its characteristics in the near-infrared atmospheric windows at 2.1 and 1.6 µm. There have been several attempts to measure the continuum in the laboratory; not only are there significant differences amongst these measurements, but there are also difficulties in extrapolating the laboratory data taken at room temperature and above to temperatures more widely relevant to the atmosphere. Validation is therefore required using field observations of the real atmosphere. There are currently no published observations in atmospheric conditions with enough water vapour to detect a continuum signal within these windows or where the self-continuum component is significant. We present observations of the near-infrared water vapour continuum from Camborne, UK, at sea level using a Sun-pointing, radiometrically calibrated Fourier transform spectrometer in the window regions between 2000 and 10 000 cm−1. Analysis of these data is challenging, particularly because of the need to remove aerosol extinction and the large uncertainties associated with such field measurements. Nevertheless, we present data that are consistent with recent laboratory datasets in the 4 and 2.1 µm windows (when extrapolated to atmospheric temperatures). These results indicate that the most recent revision (3.2) of the MT_CKD foreign continuum, versions of which are widely used in atmospheric radiation models, requires strengthening by a factor of ∼5 in the centre of the 2.1 µm window. In the higher-wavenumber window at 1.6 µm, our estimated self- and foreign-continua are significantly stronger than MT_CKD. The possible contribution of the self- and foreign-continua to our derived total continuum optical depth is estimated by using laboratory or MT_CKD values of one, to estimate the other. The obtained self-continuum shows some consistency with temperature-extrapolated laboratory data in the centres of the 4 and 2.1 µm windows. The 1.6 µm region is more sensitive to atmospheric aerosol and continuum retrievals and therefore more uncertain than the more robust results at 2.1 and 4 µm. We highlight the difficulties in observing the atmospheric continuum and make the case for additional measurements in both the laboratory and field and discuss the requirements for any future field campaign.