Atmospheric Measurement Techniques (Jul 2021)

Characterization of dark current signal measurements of the ACCDs used on board the Aeolus satellite

  • F. Weiler,
  • T. Kanitz,
  • D. Wernham,
  • M. Rennie,
  • D. Huber,
  • M. Schillinger,
  • O. Saint-Pe,
  • R. Bell,
  • T. Parrinello,
  • O. Reitebuch

DOI
https://doi.org/10.5194/amt-14-5153-2021
Journal volume & issue
Vol. 14
pp. 5153 – 5177

Abstract

Read online

Even just shortly after the successful launch of the European Space Agency satellite Aeolus in August 2018, it turned out that dark current signal anomalies of single pixels (so-called “hot pixels”) on the accumulation charge-coupled devices (ACCDs) of the Aeolus detectors detrimentally impact the quality of the aerosol and wind products, potentially leading to wind errors of up to several meters per second. This paper provides a detailed characterization of the hot pixels that occurred during the first 1.5 years in orbit. The hot pixels are classified according to their characteristics to discuss their impact on wind measurements. Furthermore, mitigation approaches for the wind retrieval are presented and potential root causes for hot pixel occurrence are discussed. The analysis of the dark current signal anomalies reveals a large variety of anomalies ranging from pixels with random telegraph signal (RTS)-like characteristics to pixels with sporadic shifts in the median dark current signal. Moreover, the results indicate that the number of hot pixels almost linearly increased during the observing period between 2 September 2018 and 20 May 2020 with 6 % of the ACCD pixels affected in total at the end of the period leading to 9.5 % at the end of the mission lifetime. This work introduces dedicated instrument calibration modes and ground processors, which allowed for a correction shortly after a hot pixel occurrence. The achieved performance with this approach avoids risky adjustments to the in-flight hardware operation. It is demonstrated that the success of the correction scheme varies depending on the characteristics of each hot pixel itself. With the herein presented categorization, it is shown that multi-level RTS pixels with high fluctuation are the biggest challenge for the hot pixel correction scheme. Despite a detailed analysis in this framework, no conclusion could be drawn about the root cause of the hot pixel issue.