Verifying Snow Photochemistry Models by Combining Actinometry and UV Radiometry

In the past few years, experiments and modeling studies have shown that chemical processes in the snow pack may have significant impacts on the chemistry of the atmosphere. A number of these studies have suggested that solar UV radiation penetrating the snow pack is the driving force for some of these chemical processes. Owing to the need to quantify the extent of the photochemical processing in the snow pack, radiative transfer models have been developed for use in the snow pack.

The goal of this study is the development and performance testing of radiative transfer models constrained with photometric measurements of UV irradiance. To test these models, we compare the modeled photolysis rates to photolysis rates measured by chemical actinometry within an artificial snow-like scattering medium. This snow analog was made by flash freezing small droplets of a dilute actinometer solution in liquid nitrogen. Radiative transfer models based on both the delta-Eddington and the DISORT methods were used. The DISORT model was TUV, developed by Lee-Taylor and Madronich. To use these models, measurements of the snow analog's optical properties are needed. The light attenuation and intensity were measured in the scattering medium using a cosine- response sensor coupled to a spectrometer by fiber optic. With these optical properties constraining the model, we can then predict photolysis rates as a function of depth within the snow analog. The veracity of the radiative transfer models in the absence of light scattering was tested using 2-nitrobenzaldehyde (NBA) as an actinometer in liquid solution. The model results were found to have good agreement with the actinometer in this control experiment. The photolysis rate of NBA with respect to depth was measured from the conversion of the NBA to 2-nitrosobenzoic acid. The snow analog was irradiated with a UV light source. After exposure, samples of the snow analog were collected with respect to depth and the conversion of the actinometer was determined by measuring the UV absorption spectra of the melted samples. The resulting measurements of the photolysis rate of NBA were compared to the model derived values to assess the performance of the models.