Air pollution has a significant detrimental effect on human/ecosystem health and is an important issue for a rapidly urbanising world. Anthropogenic NOx (NO and NO2) emissions have a range of health and environmental impacts. Reliable measurements of NOx are used to inform a range of science and policy goals. Satellite observations of tropospheric NO2 give a synoptic view of global and regional air pollution and are essential to constraining NOx emission estimates. However, coarse spatio-temporal resolution limits their utility to analyse urban pollution. The Compact Air Quality Spectrometer (CompAQS), a UV-VIS hyperspectral instrument, was developed to provide high spatial resolution imaging of urban pollution. The work described in this thesis covers the development of alternative retrieval algorithms of tropospheric NO2 for CompAQS. A variant on the Differential Optical Absorption Spectroscopy (DOAS) technique was developed using unpolluted Earthshine reference spectra, and was tested using modelled and operational satellite spectra. Tropospheric columns retrieved with this technique agreed well with operational products (r2 = 0:85), particularly over heavily polluted regions (r2 = 0:96). A neural network retrieval employing only several wavelengths was also developed and tested on modelled spectra, which demonstrated low fitting errors ( < 10%) even with noisy data. Air mass factors were calculated to correct the data for other influences (e.g. surface albedo) for a CompAQS demonstrator test flight campaign over Leicester in February 2013. While CompAQS could resolve high-resolution pollution features, the DOAS fit uncertainty dominated the retrieval error (37%), indicating that better instrument calibration would be required to improve its eventual use as a satellite instrument. The retrieval algorithms demonstrated in this work could be tailored to support satellite adaptations of CompAQS. However, further research investigating the reliability and resolution of assumed a priori information (e.g. the NO2 vertical profile) will be required to produce a full mission study concept.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:677422 |
| Date | January 2015 |
| Creators | Anand, Jasdeep Singh |
| Contributors | Leigh, Roland ; Monks, Paul |
| Publisher | University of Leicester |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/2381/36126 |
Page generated in 0.0018 seconds