Atmospheric trace gas retrievals and monitoring using a medium-resolution spectroradiometer

Accurate retrieval of atmospheric carbon dioxide from ground-based solar spectral measurements in the near-infrared spectral range requires precise characterisation of interfering absorbers, such as atmospheric water vapour and aerosols. In the near-infrared spectral region, water vapour absorption strongly overlaps with carbon dioxide (CO₂) features, which is making reliable carbon dioxide retrieval highly sensitive to uncertainties in the integrated water vapour (IWV) column.

The primary objective of this work is the development of a robust CO₂ retrieval algorithm using compact spectroradiometers with moderate spectral resolution; however, as a critical prerequisite, an accurate water vapour retrieval framework must be established.

In this study, IWV is retrieved from direct solar spectral irradiance measured by a compact Bi-Tec Sensor (BTS) spectroradiometer using an optimised wavelength selection approach. The optimum wavelengths were identified through synthetic radiative transfer simulations covering a wide range of atmospheric conditions at Davos, Switzerland. Wavelengths exhibiting strong water vapour sensitivity and high stability across seasons were selected and subsequently applied to real atmospheric measurements. The resulting BTS-derived IWV algorithm was evaluated against co-located GPS IWV (AGNES) observations and compared with AERONET IWV retrievals.

The formulated BTS IWV retrieval algorithm exhibits good agreement with GPS (AGNES) measurements, with a mean bias of -1.01 mm and a standard deviation of 0.80 mm over the analysed period. In comparison, the co-located AERONET IWV retrieval shows a larger mean bias of -2.60 mm and higher residual variability of 1.82 mm, indicating the BTS-based algorithm improved stability and accuracy of water vapour retrieval. These results demonstrate that the careful selection of physically meaningful and spectrally stable wavelengths identified through synthetic radiative-transfer modelling and subsequently applied to real atmospheric measurements leads to a substantial improvement in retrieval performance.

Building on this validated water vapour retrieval, ongoing work focuses on integrating the IWV product into a CO₂ retrieval framework in the 1.6 µm to 2 µm spectral region. The established IWV retrieval provides a critical constraint for reducing systematic errors and improving the robustness of CO₂ estimation from compact spectroradiometers.