Logo Logo
Hilfe
Kontakt
Switch language to English
Towards CO2 emission monitoring with passive air- and space-borne sensors
Towards CO2 emission monitoring with passive air- and space-borne sensors
Coal-fueled power plants are responsible for 30 % of anthropogenic carbon dioxide (CO2) emissions and can therefore be considered important drivers of climate warming. The 2015 Paris Climate Accord has established a global stock take mechanism, which will assess the progress of global carbon emission reduction policies in five-yearly tallies of worldwide emissions. However, there exists no independent monitoring network, which could verify such stock takes. Remote sensing of atmospheric CO2 concentrations from air- and space-borne sensors could provide the means of monitoring localized carbon sources, if their ground sampling distance is sufficiently fine (i.e. below the kilometer scale). Increased spatial resolution can be achieved at the expense of decreasing the spectral resolution of the instrument, which in turn complicates CO2 retrieval techniques due to the reduced information content of the spectra. The present thesis aims to add to the methodology of remote CO2 monitoring approaches by studying the compromise between spectral and spatial resolution with CO2 retrievals from three different sensors. First, the trade-off between coarse spectral resolution and retrieval performance is discussed for a hypothetical imaging spectrometer which could reach a spatial resolution of ~50×50 m2 by measuring backscattered sunlight in the short wave infrared spectral range at a resolution of ∆λ ~ 1 nm. To this end, measurements of the Greenhouse gases Observing SATellite (GOSAT) at ∆λ = 0.1 nm are artificially degraded to coarser spectral resolutions to emulate the proposed sensor. CO2 column retrievals are carried out with the native and degraded spectra and the results are compared with each other, while data from the ground based Total Carbon Column Observing Network (TCCON) serve as independent reference data. This study identifies suitable retrieval windows in the short wave infrared spectral range and a favorable spectral resolution for a CO2 monitoring mission. Second, CO2 column retrievals are carried out with measurements of the air-borne AVIRIS-NG sensor at a spectral resolution of ∆λ = 5 nm. This case study identifies advantageous CO2 retrieval configurations, which minimize correlations between retrieval parameters, near two coal-fired power plants. A bias correction method is proposed for the retrievals and a plume mask is applied to the retrieved CO2 enhancements to separate the CO2 emission signal from the atmospheric background. Emission rates of the two facilities are calculated under consideration of the local wind speed, compared to a public inventory and discussed in terms of their uncertainties. Third, CO2 retrievals are extended to spectral resolutions on the order of ∆λ ~ 10 nm by analyzing spectra of the specMACS imager near a small power plant. Retrieval effects that hamper the detection of the source signal are discussed.
Not available
Wilzewski, Jonas S.
2021
Englisch
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Wilzewski, Jonas S. (2021): Towards CO2 emission monitoring with passive air- and space-borne sensors. Dissertation, LMU München: Fakultät für Physik
[thumbnail of Wilzewski_Jonas_Simon.pdf]
Vorschau
PDF
Wilzewski_Jonas_Simon.pdf

17MB

Abstract

Coal-fueled power plants are responsible for 30 % of anthropogenic carbon dioxide (CO2) emissions and can therefore be considered important drivers of climate warming. The 2015 Paris Climate Accord has established a global stock take mechanism, which will assess the progress of global carbon emission reduction policies in five-yearly tallies of worldwide emissions. However, there exists no independent monitoring network, which could verify such stock takes. Remote sensing of atmospheric CO2 concentrations from air- and space-borne sensors could provide the means of monitoring localized carbon sources, if their ground sampling distance is sufficiently fine (i.e. below the kilometer scale). Increased spatial resolution can be achieved at the expense of decreasing the spectral resolution of the instrument, which in turn complicates CO2 retrieval techniques due to the reduced information content of the spectra. The present thesis aims to add to the methodology of remote CO2 monitoring approaches by studying the compromise between spectral and spatial resolution with CO2 retrievals from three different sensors. First, the trade-off between coarse spectral resolution and retrieval performance is discussed for a hypothetical imaging spectrometer which could reach a spatial resolution of ~50×50 m2 by measuring backscattered sunlight in the short wave infrared spectral range at a resolution of ∆λ ~ 1 nm. To this end, measurements of the Greenhouse gases Observing SATellite (GOSAT) at ∆λ = 0.1 nm are artificially degraded to coarser spectral resolutions to emulate the proposed sensor. CO2 column retrievals are carried out with the native and degraded spectra and the results are compared with each other, while data from the ground based Total Carbon Column Observing Network (TCCON) serve as independent reference data. This study identifies suitable retrieval windows in the short wave infrared spectral range and a favorable spectral resolution for a CO2 monitoring mission. Second, CO2 column retrievals are carried out with measurements of the air-borne AVIRIS-NG sensor at a spectral resolution of ∆λ = 5 nm. This case study identifies advantageous CO2 retrieval configurations, which minimize correlations between retrieval parameters, near two coal-fired power plants. A bias correction method is proposed for the retrievals and a plume mask is applied to the retrieved CO2 enhancements to separate the CO2 emission signal from the atmospheric background. Emission rates of the two facilities are calculated under consideration of the local wind speed, compared to a public inventory and discussed in terms of their uncertainties. Third, CO2 retrievals are extended to spectral resolutions on the order of ∆λ ~ 10 nm by analyzing spectra of the specMACS imager near a small power plant. Retrieval effects that hamper the detection of the source signal are discussed.