Theoretical Approach to the Radiance-to-Flux Conversion in the EarthCARE Mission Framework.
The research of the Thesis, Theoretical Approach to the Radiance-to-Flux Conversion in the EarthCARE Framework, is aimed at studying the instantaneous TOA radiance-to-flux conversion for the prospective Broad-Band Radiometer (BBR) on-board the EarthCARE (Earth Clouds Aerosols and Radiation Explorer) platform, through the development of theoretical angular distribution models based on the specific designing features of the instrument. The inversion procedure has been undertaken to obtain the apparent TOA fluxes according to the flux retrieval accuracy required for the mission. The ADMs employed in previous missions make use of a single radiance in the inversion process. With the specific BBR along-track configuration, three simultaneous radiances point the same target, thus providing more information on the anisotropy of the observed scene. This additional information is employed to construct more accurate inversion schemes than those using a single radiance, thus the objective has been to construct theoretical ADMs based on the specific characteristics designed for this pioneering instrument. However, the resulting angular models are not a finished product to be directly applied over EarthCARE data. This is an attempt to study one potential solution and assess its drawbacks and advantages.
Considering that the use of existing ADMs have important shortcomings, the Thesis opens an unexplored study line by which detailed numerical calculations are used to assess and construct ADMs. A specific module of the EarthCARE Simulator has been employed to carry out the broad-band radiative transfer computations, namely the semi-analytical Monte-Carlo radiative transfer code.
In spite of the difficulty associated with addressing the unlimited atmospheric conditions needed in order to determine ADMs based on model calculations, a theoretical database has been constructed considering different surface, atmospheric and cloud conditions. The variations on these parameters are conditioned to the EarthCARE satellite orbit and physical ligatures. At the end, more than 8x104 detailed scenes has been produced, building up significant ensembles with a robust statistical analysis, and including realistic 3-D cloud effects. A high number of calculations has been performed on the Computing Centre from the University of Valencia while the European Space Agency (ESA) facilitated a significant computing resource by allowing us to implement the EarthCARE Simulator on the ESA Earth Observation Grid Processing-on-Demand (EO G-POD).
Three different inversion methodologies have been specifically developed to take advantage of the along-track scanner configuration of the EarthCARE BBR. Flux can be estimated, as a result of the study, by using an optimized classical inversion procedure with different effective radiance definitions, namely along-track and average methods, or defining the flux as the linear combination of the three-view radiances. These conversion methodologies has been applied to a BBR-like CERES representative dataset, and the results has been compared to the flux estimates obtained by the single-view based CERES ADMs. It has been shown that the application of theoretical ADMs on satellite-measured radiances is far from straightforward. The current approach is presently far from complete and requires significant further research. Empirical ADMs must farther studied before they are rule out.
A number of multi-layer bins have been generated for a prototype BBR ADM scene definition, still experimental though. Twenty five multi-layered realistic cloud scenes have been built for this study employing a 3-D stochastic cloud model generator. Since the cloud-imager retrievals are insufficient for the BBR scene identification (ID) in multi-layer cloud, the radiance-to-flux conversion provides large errors. The use of the active instruments helps to determine the type of clouds involved in the scene in most multi-layer situations. The Thesis demonstrates the potential improvement of the radiance-to-flux conversion, by adequately selecting the ADM scene definition and including information from active instruments on-board EarthCARE platform in the scene ID.
Advisor:Donovan, David P.; López Baeza, Ernesto
School:Universitat de València
Source Type:Master's Thesis
Date of Publication:06/02/2008