Harmonization of space-borne infra-red sensors measuring sea surface temperature

Merchant, C. J. ORCID: https://orcid.org/0000-0003-4687-9850, Block, T., Corlett, G. K., Embury, O. ORCID: https://orcid.org/0000-0002-1661-7828, Mittaz, J. P. D. and Mollard, J. D. P. (2020) Harmonization of space-borne infra-red sensors measuring sea surface temperature. Remote Sensing, 12 (6). 1048. ISSN 2072-4292 doi: 10.3390/rs12061048

Abstract/Summary

Sea surface temperature (SST) is observed by a constellation of sensors, and SST retrievals are commonly combined into gridded SST analyses and climate data records (CDRs). Differential biases between SSTs from different sensors cause errors in such products, including feature artefacts. We introduce a new method for reducing differential biases across the SST constellation, by reconciling the brightness temperature (BT) calibration and SST retrieval parameters between sensors. We use the Advanced Along-Track Scanning Radiometer (AATSR) and the Sea and Land Surface Temperature Radiometer (SLSTR) as reference sensors, and the Advanced Very High Resolution Radiometer (AVHRR) of the MetOp-A mission to bridge the gap between these references. Observations across a range of AVHRR zenith angles are matched with dual-view three-channel skin SST retrievals from the AATSR and SLSTR. These skin SSTs act as the harmonization reference for AVHRR retrievals by optimal estimation (OE). Parameters for the harmonized AVHRR OE are iteratively determined, including BT bias corrections and observation error covariance matrices as functions of water-vapor path. The OE SSTs obtained from AVHRR are shown to be closely consistent with the reference sensor SSTs. Independent validation against drifting buoy SSTs shows that the AVHRR OE retrieval is stable across the reference-sensor gap. We discuss that this method is suitable to improve consistency across the whole constellation of SST sensors. The approach will help stabilize and reduce errors in future SST CDRs, as well as having application to other domains of remote sensing.

Item Type	Article
URI	https://reading-clone.eprints-hosting.org/id/eprint/89664
Item Type	Article
Refereed	Yes
Divisions	Science > School of Mathematical, Physical and Computational Sciences > National Centre for Earth Observation (NCEO) Science > School of Mathematical, Physical and Computational Sciences > Department of Meteorology
Publisher	MDPI
Download/View statistics	View download statistics for this item