Climate change and atmospheric pollution are current topical issues, given their possible dramatic effects from the health, social and economical points of view. Ozone is one of the most important trace gases in the Earth’s atmosphere, since, despite of its small amount, it prevents harmful solar radiation from reaching the surface. Human activities have a recognized role on stratospheric ozone depletion, hence a possible increase of surface UV radiation, and may contribute to enhance tropospheric ozone concentration. In this thesis, novel approaches to the inversion of nadir UV/VIS satellite Earth’s radiance spectra for the retrieval of height-resolved ozone information are presented. The considered platforms are ESA EnviSat-SCIAMACHY and NASA-Aura OMI, which are particularly suited for these tasks given their combined relatively high spectral and spatial resolutions. Both ozone concentration profiles and tropospheric ozone columns are retrieved by Neural Network algorithms, which were proven to be robust on systematic errors and calibration uncertainties. They are likely to work better than Optimal Estimation with respect to cloudy scenarios or in presence of significant aerosols burdens and, once trained, perform retrievals in real time. The information content of the VIS band is critically discussed. Design issues are presented, as well as results both global and at mid-latitudes. NNs were also considered for retrieving other atmospheric parameters, i.e. temperature profiles and nitrogen dioxide columns; the underlying idea here is to provide an integrated tool for a more complete characterization of the atmospheric state from the same input vector. Finally, an approach to the validation of NNs algorithms in a GRID environment over extensive datasets, is presented and briefly discussed.
Sellitto, P. (2009). Neural networks algorithms for the estimation of atmospheric ozone from Envisat-SCIAMACHY and Aura-OMI measurements.
|Titolo:||Neural networks algorithms for the estimation of atmospheric ozone from Envisat-SCIAMACHY and Aura-OMI measurements|
|Data di pubblicazione:||6-mag-2009|
|Anno Accademico:||A.A. 2008/2009|
|Corso di dottorato:||Geoinformazione|
|Settore Scientifico Disciplinare:||Settore ING-INF/06 - Bioingegneria Elettronica e Informatica|
|Tipologia:||Tesi di dottorato|
|Citazione:||Sellitto, P. (2009). Neural networks algorithms for the estimation of atmospheric ozone from Envisat-SCIAMACHY and Aura-OMI measurements.|
|Appare nelle tipologie:||07 - Tesi di dottorato|