Tomographic reconstruction of gas plumes using scanning DOAS

This paper presents a method to reconstruct the gas distribution inside a vertical cross section of a gas plume by combining data from two or more scanning DOAS instruments using a tomographic algorithm. The method can be applied to gas plumes from any single, elevated point source, such as a volcano or industrial chimney. Such two-dimensional concentration distributions may prove to be useful for example in plume chemistry, dispersion and environmental impact studies. Here we show the case with one scanning DOAS instrument located on each side of the plume, which is the easiest and most economic setup as well as the most useful in routine monitoring of e.g. volcanic gas emissions. The paper investigates the conditions under which tomographic reconstructions can be performed and discusses limitations of this setup. The proposed method has been studied theoretically by numerical simulations and has been experimentally tested during two field campaigns, with measurements of SO2 emissions from a volcano and a power plant. The simulations show that, under good measurement conditions, the algorithm presented performs well, which is further confirmed by the experimental results.