Day 4

Detailed paper information

Back to list

Paper title A Metrological Approach to Uncertainty Analysis for the TRUTHS Satellite Mission
Authors
  1. Jacob Fahy National Physical Laboratory Speaker
  2. Samuel E. Hunt National Physical Laboratory NPL
  3. Nigel Fox National Physical Laboratory
  4. Paul Green National Physical Laboratory, UK
  5. Yoshiro Yamada National Physical Laboratory, UK
Form of presentation Poster
Topics
  • B1. Calibration, validation and data quality, FRM
    • B1.01 SI-Traceable Satellites - a Gold Standard for Climate and Intercalibration
Abstract text TRUTHS (Traceable Radiometry Underpinning Terrestrial- and Helio-Studies) is a UKSA-led climate mission that is in development as part of ESA Earth Watch programme, which has the aim of establishing of high accuracy SI-traceability on-orbit to improve estimates of the Earth’s radiation budget and other parameters by up to an order of magnitude. The high-accuracy that its SI-traceable calibration system enables (target uncertainty of 0.3 % (k=2)) allows TRUTHS observations to be used both directly as a climate benchmark and as a reference sensor for upgrading the calibration of other sensors on-orbit.

In order to the assess the proposed instrument design against the strict uncertainty requirements, a rigorous and transparent evidence-based uncertainty analysis is required. This paper describes a metrological analysis of the radiometric processing of the TRUTHS L1b products derived from Top of the atmosphere observed photons, including an analysis of the On-Board Calibration System (OBCS) performance. At the heart of the OBCS is the Cryogenic Solar Absolute Radiometer (CSAR) which provide the primary traceability to SI. The OBCS mirrors concepts used in national standards laboratories for measurement of optical power and spectral radiance and irradiance and in TRUTHS links the calibration of the Hyperspectral Imaging Spectrometer (HIS) to SI.

The analysis follows the framework outlined in the EU H2020 FIDelity and Uncertainty in Climate data records from Earth Observations (FIDUCEO) project, which uses a rigorous GUM-based approach to provide uncertainties for Earth Observation products and contains a number of documentational and visualisation concepts that aid interrogation and interpretation. Initially the measurement functions for each instrument on board TRUTHS are defined, and a corresponding ‘uncertainty tree diagram’ visualisation produced. From this, error effects are identified and described using ‘effects tables’, which document the associated uncertainty, sensitivity coefficients and error-correlation structure, providing the necessary information to propagate the uncertainty to the final product. Combining this uncertainty information allows for the total uncertainty of a quantity (e.g. radiance) to be estimated, at a per-pixel level, which can then be analysed based on the source of the uncertainty or its error-correlation structure (e.g., random, systematic, etc.).

An extension of this analysis is End-to-End Metrological Simulator (E2EMS) for the TRUTHS L1b Products, adding both a forward model of the TRUTHS sensor (input radiance to measured counts), and a calibration model (measured counts to calibrated radiance), to understand the product quality an contributions from the proposed schemes and algorithms.