Researchers at the National Oceanography Centre Southampton, led by Professor Ian Robinson, have pioneered a new approach to measuring the temperature of the surface of the sea by combining data from different satellites and comparing them with information from ship-borne sensors, known as radiometers. The resulting maps of sea surface temperature (SST) have led to better weather prediction and marine forecasting for improved safety at sea, and underpin more reliable monitoring of the effects of climate change on the ocean.
From 1991, the European Space Agency (ESA) started to use improved satellite systems to detect changes in global SST by measuring the emitted infrared radiation. The Along-track Scanning Radiometer (ATSR) system aimed to produce SST images with accuracy better than 0.2 degree at a resolution of one kilometre. These precise measurements are needed for accurate weather forecasting and climate change monitoring, but the several days needed to build up global coverage at first prevented their use for short-term forecasting.
Moreover, for such high quality satellite data to be reliable, they must be independently checked against observations made at sea level. Conventional thermometers on buoys measure the temperature a metre or more below the sea surface, but this can differ considerably from the true sea surface in contact with the atmosphere and viewed by satellites. The sun warms the surface while wind and cloud shadows have a cooling effect, causing variations of up to five degrees between the true skin and what the buoy measures. This creates large uncertainty when validating satellite SST data using buoy temperatures.
As part of his work, Professor Robinson helped bring together scientists from research organisations around the world to pool their knowledge, in the interests of creating ever more accurate SST maps for everyone to use. In the early 2000s, within the Group for High Resolution Sea Surface Temperature (GHRSST), researchers worked directly with those producing data from different space agencies, in order to understand the reasons why different satellite systems produce differing observations.
They proposed a new approach which would allow the different SST datasets to be blended, thus allowing the strengths of one system to offset the weaknesses of another. To make this possible, they reached agreement about consistent data formats, about the need for independent assessment of the accuracy of each different dataset, and the acquisition of environmental information such as wind speed at each time and place where a satellite measurement was acquired.
Professor Robinson and his colleague Dr Craig Donlon, who now works at the European Space Agency, achieved two important breakthroughs which contributed strongly to the success of the GHRSST approach and its adoption by international consensus for all major global SST data products. Firstly, starting in 2004 through the ESA-sponsored Medspiration project, they designed and constructed the prototype GHRSST processing system for SST data from several satellites. For the first time it became possible to compare ATSR data with those from US, Japanese and other European sensors, confirming its superior stability and accuracy. By 2006 ATSR data were being used as a reference for bias correcting all other satellite inputs to a new blended SST analysis developed by the UK Met Office using the Medspiration data. By this means the superior accuracy of ATSR data was spread to systems with better global coverage so that maps of high resolution, high accuracy SST could be updated daily for operational use.
Secondly, over ten years Robinson and Donlon developed a new generation of radiometers which can be fixed to ships to take the temperature of the topmost millimetre of the sea. In 1996, Southampton had secured European Union funding to draw together other SST experts from organisations such as ESA, NASA, Eumetsat and the UK and French Met Offices, to explore the accuracy of radiometry for measuring SST. Building on this, they designed and built the first infrared sea surface temperature autonomous shipborne radiometer (ISAR); it can measure SST to accuracies of better than 0.1 degree. Since 2004, ISARs have been in continuous, unattended use on ferries between the UK and Spain and its results have enabled independent, traceable verification of the ATSR satellite data.
Today, the GHRSST processing model, prototyped and tested by the Medspiration project at Southampton has been adopted by operational agencies and is at the core of most satellite SST products produced around the world. In 2007, a rapid melting of the Arctic icecap caused problems for the UK Met Office weather forecasts, which were resolved by switching from their conventional SST data to the new high resolution analysis incorporating ATSR data and developed as a direct spin-off from the Medspiration Project at Southampton.
So far ten ISAR systems have been built at Southampton and they are now being used around the world by ocean and meteorological institutes in the USA, Japan, China and Denmark and by the Royal Navy. Their results have revolutionised the way satellite SST data are validated, providing traceable accuracy for detailed temperature maps that are now in widespread use for forecasting and climate monitoring.
Insight from Southampton researchers into measuring SST have resulted in pioneering new systems that have improved forecasting accuracy across the globe. It has increased our understanding of climate change and improved safety for seafarers.
Southampton researchers are now involved in the validation of a soon to be launched ESA satellite sensor to replace ATSR. They are also considering ways of adapting ISAR to measure the air temperature at the sea surface to better understand its influence on ocean temperatures in climate change.
|Evaluation of accuracy and uncertainty of EO parameters for global air-sea fluxes||Active||Other|