Cloud optical thickness retrievals in the Arctic from zenith-looking spectral measurements.

Clouds are one of the Earth system components which need to be comprehensively understood and from which most of the uncertainties in climate models and forecasts originate. Cloud Optical Thickness (COT) is of utmost importance for estimating the effects of clouds on the upwelling and downwelling radiation. Still, few issues arise when retrieving it over high albedo surfaces from ground-based or satellite measurements. We used a mixed approach that combines radiative transfer model simulations (libRadtran v. 2.0.4) with ground-based zenith spectral measurements (310--950 nm) at the Thule High Arctic Atmospheric Observatory (THAAO $76.5^{\circ}$N, $68.8^{\circ}$W, Greenland) to estimate COT. We present the theoretical background and the application of different algorithms for COT retrievals in low- and high-surface albedo conditions for liquid water clouds and homogeneous cloud cover. We compare various spectral algorithms for COT with retrievals relying on the instruments available at THAAO. In the case of high-surface albedo, we find good agreement among the methods when using spectral algorithms that include external information on the cloud base height.