A daily, global Sea Surface Temperature (SST) data set is produced at 1-km (also known as ultra-high resolution) by the JPL ROMS (Regional Ocean Modeling System) group. Based on the level 2 preprocessed (L2P) data products from the Global High-Resolution Sea Surface Temperature (SST) Pilot Project (GHRSST-PP), the input SST data include infrared (IR) sensors (e.g., Advanced Very High Resolution Radiometer or AVHRR, Moderate Resolution Imaging Spectroradiometer or MODIS, Advanced Along-Track Scanning Radiometer or AATSR) with a spatial resolution of 1~2 km, the Geostationary Environmental Satellites (GOES, METOP) with a spatial resolution of 6 km, and microwave sensors (e.g., Advanced Microwave Scanning Radiometer Earth Observing System or AMSR-E, Tropical Rainfall Measuring Mission Microwave Imager or TMI) with a spatial resolution of 25 km. The in-situ SSTs obtained from the Global Ocean Data Assimilation Experiment (GODAE) server consisting of thousands of daily ship and buoy SST measurements are also used. The goal is to provide a global optimal estimate of SST at the highest possible spatial resolution.
In order to combine all the available SST data sets at various spatial resolutions, we have developed a multi-scale two-dimensional variational (MS-2DVAR) blending algorithm. This MS-2DVAR algorithm is characterized by inhomogeneous and anisotropic background error covariances, which are of particular importance for coastal oceans (Chao, Y., Z. Li, J. D. Farrara, and P. Huang: Blended sea surface temperatures from multiple satellites and in-situ observations for coastal oceans. Journal of Atmospheric and Oceanic Technology, 26 (7), 1435-1446, 10.1175/2009JTECHO592.1, 2009).
We provide images for the global ocean as well as six selected regions: U.S. west coast, U.S. East Coast, Gulf of Mexico and Caribbean Sea, Hawaii Islands, Peru Coastal Ocean, and South China Sea. Users can generate on-demand images with the appropriate color scale over any region of interests by specifying the starting and ending longitudes and latitudes.
The Global 1-km SST data are produced daily in near real-time: yesterday's image and data are usually available in the morning. The version 1.0 G1SST data product started from September 1, 2008. The version 2.0 G1SST data product starts from May 1, 2009, and includes more input data, including METOP, AATSR, and In-Situ. The Google Earth API interface and the OpenDAP/THREDDS data server are also implemented.
We are in the process of conducting a systematic evaluation of the G1SST blended SST product. Improvements in our blending algorithm to take into account issues such as the diurnal cycle, skin vs bulk SST, a better representation of the input data error and error covariances are also planned. Depending upon the computational resources and the availability of the GHRSST L2P data products, we will generate a longer time series of SST prior to September 1, 2008.
We welcome comments and feedback from users, and look for potential collaborations with researchers to further evaluate and improve our G1SST product.
The JPL G1SST Team: Zhijin Li, Peggy Li, Benyang Tang, and Quoc Vu