Introduction, purpose and scope

SST is an essential climate variable [Bojinski et al., 2014] that is crucial for a wide range of applications, such as e.g. input to NWP models [Brasnett and Colan, 2016, Chelton and Wentz, 2005] and ocean models [Le Traon et al., 2015, Liang et al., 2017], for understanding air-sea interactions [Monzikova et al., 2017, Ning et al., 2018] and monitoring the climate [Merchant et al., 2019]. Observations of SST from IR satellites have been available since the early 1980’s, however these observations are affected by atmospheric aerosols and limited to clear-sky conditions [Merchant et al., 2019, Reynolds et al., 2002, Vázquez-Cuervo et al., 2004]. PMW observations of SST are an important and complementary alternative as they are not affected by non-precipitating clouds and the effect from aerosols is negligible [Donlon et al., 2007, Ulaby et al., 1981, Wentz et al., 2000].

The first PMW SST observations were retrieved in 1978 from measurements by SMMR, which was carried on Seasat 1 and Nimbus 7 [Lipes, 1982, Milman and Wilheit, 1985]. However, the retrieved SSTs contained large errors [Milman and Wilheit, 1985] and were of a very coarse spatial resolution, which limited their usefulness. In 1997, the TMI was launched, providing SST observations between 40\(^{\circ}\)N/S in the period 1998 to 2014 [Kummerow et al., 1998, Wentz, 2015]. Global and accurate PMW SSTs were first measured in 2002 by AMSR-E, from NASA’s Aqua platform [Kawanishi et al., 2003]. In October 2011, AMSR-E ceased normal operations and in 2012 the follow-on mission, AMSR2, was launched onboard GCOM-W1 [Imaoka et al., 2010]

The resolution of the current microwave imagers are not enough to capture subscale to mesoscale variability. Furthermore, they suffer from coastal and sea ice contamination at the 6.9 GHz channels due to the large field of view (e.g. AMSR2 has a 64 \(\times\) 32 km field of view at 6.9 GHz). As the spatial resolution of SST observations are limited by e.g. the ratio of the wavelength of the measurement to the antenna diameter [Wentz and Meissner, 2000], a larger antenna is necessary to obtain a high spatial resolution. Therefore, an improvement of the spatial resolution of measurements from the 6.9 and 10.7 GHz channels could result in a significant improvement of PMW SST observations and their information content in global products and regional analysis systems.

CIMR is currently being prepared by ESA as a part of the Copernicus Expansion program of the European Union (http://www.cimr.eu/). CIMR is designed to observe high-resolution and high-accuracy global PMW measurements of SST, WS, SSS and SIC as primary variables from radiometer channels at 1.4, 6.9, 10.7, 18.7 and 36.5 GHz from a Sun-synchronous polar orbit [Donlon, 2020].

This ATBD describes the algorithm for the retrieval of SST from CIMR. The algorithm consists of a statistically-based regression retrieval for SST and an optional regression-based retrieval for WS.