Methodology for determination of ozone vertical distribution elements in the atmosphere from IKFS-2 measurements

Ozone is crucial for the chemical and thermal balance of the atmosphere. Due to different lifetime, the formation and destruction mechanisms of ozone, as well as different trends of its variability and the role of ozone in different atmospheric layers, it is important to obtain information not only on the total ozone column, but also on its spatial (vertical and horizontal) and temporal distribution. To obtain information on the ozone vertical distribution, the method of outgoing thermal radiation can be used. The equipment of the Meteor-M No. 2 series Russian meteorological satellite includes the hyperspectral infrared sounder IKFS-2, which measures the outgoing IR thermal radiation in the range of 660-2000 cm^-1. The information content estimates of the IKFS-2 measurements show that the mean number of degrees of freedom for the signal (DOFS) with respect to the vertical ozone profile depends on latitude/season and is close to 4 on average.

We present an algorithm for solving the inverse problem using a physical and mathematical approach based on the optimal estimation method. To consider the uncertainties caused by the forward problem parameter errors (temperature and H2O profiles, surface temperature, etc.) and to determine the optimal algorithm for solving the inverse problem for ozone profile retrievals, we carried out closed-loop numerical experiments. On average for all latitudes, the estimate of the algorithm error with respect to the total ozone column is 2.3 %. Measurements of total ozone column at mid-latitudes have the highest accuracy (the error is estimated as 1.7 %). In the future, uncertainties in the emissivity should be taken into account.

Numerical experiments show that in the troposphere the mean error in determining the ozone content from IKFS-2 measurements with no smoothing error is 5-10 % depending on latitude (the smallest error occurs at high latitudes). The error in determining stratospheric ozone content is 5-15 % (the smallest error occurs at low latitudes). The largest error with respect to the ozone profile is observed in the UTLS region, and is 20-30 %.

We processed the IKFS-2 measurements. The results were validated with vertical profiles by ozonesondes. Mean differences show that, on average, the IKFS-2 algorithm overestimates O3 by 10-25 % in the UTLS region, while it underestimates O3 by ∼5-10 % in the stratosphere.

The authors acknowledge Saint-Petersburg State University for a research grant 124032000025-1.