The average British adult spends 11 minutes a week checking the weather forecast either on TV, a phone app or the internet, according to a pre-Covid-19 poll of 2,000 people (opens in a new tab), commissioned by Bristol Airport. We have all become familiar with satellite images of the atmosphere and surface of the Earth which have revolutionised the way weather updates are shared, with TV presenters regularly discussing weather patterns revealed by satellite data.
Over the last 40 years, the same data have led to large improvements in the ability of computer models – used to predict the weather based on physical equations and an estimate of current conditions – to forecast weather and seasonal climate over days, weeks and even months ahead. Improved satellite data helps us improve our estimate of the current weather, which leads to improved weather forecasts. Similarly, longer records of climate observations help us better test our climate models, which are similar to weather models but run over longer time periods, and better understand how climate works and how it is changing. The longer the period for which we have continuous, consistent data measuring the Earth system, the more confidence we will have in this testing and understanding.
These long satellite-derived climate data sets are used, and will continue to be used, by current and future generations of scientists to monitor and understand our changing climate.
Important role of Sentinel-6 satellite mission
Historical datasets, made up of data carefully combined from different satellite missions, can be used to check on the quality of longer-term climate models and help us improve them. It’s important that these models work at the forefront of scientific understanding because they are increasingly important for informing climate action.
Investment now in the climate satellite missions of tomorrow, such as the recent Sentinel-6 mission (opens in a new tab) which will gather sea level (altimetry) data, provides both real-time data for weather forecasters and additional data for historical climate records, supplying vital information for future generations of scientists to be able to continue to understand and monitor Earth’s climate.
Rainfall underestimate revealed by taking the sea’s temperature from space
Recently, I was part of a team, led by Dr Peter Good at the Met Office, using satellite datasets to quality-check global climate models. We analysed sea surface temperature, rainfall and surface wind data and discovered that the global climate models are underestimating the increase of rainfall in tropical locations that accompanies a warmer local sea surface temperature when simulating and predicting tropical rainfall patterns.
Our resulting study, published in Nature (opens in a new tab), found that important low-altitude wind patterns in the wider tropical region were stronger than most models simulate and this was a potential cause of discrepancies. Improving these deficiencies could lead to more confidence in model projections of future changes in rainfall and wind patterns due to climate change, vital for hundreds of millions of people in the tropics who rely on rain-fed agriculture or are vulnerable to flooding.
Improvements could also increase the accuracy of current weather forecasting, especially on seasonal time scales. Now that we’ve identified the issue, our scientific colleagues who refine and develop climate models (and weather forecasting models) can hopefully integrate this new scientific understanding into the models.
New missions are securing scientific capability for future generations
For studies of physical processes in the current climate, including studies investigating how climate has been changing in the recent past, we need long and consistent data records. For satellite instruments, which can ‘drift’ in their accuracy over time and which can differ from satellite to satellite, we also need overlaps in time between old and new instruments so we can calibrate their measurements and combine them into long reliable datasets.
These are already useful now, but the data we produce today will be even more useful to climate scientists in the future who will have longer records and (hopefully) better analysis methods and computer models.
One thing those future scientists will not be able to do, however, is travel back in time to launch more satellites, so we must invest now in our Earth observation system. That’s why the launch of satellites like Sentinel-6, is so important.
Climate satellite data records used in this study
Satellite observations over multiple decades were vital for this study.
First, the sea surface temperature products combine direct measurements (from ships and buoys) with satellite data (which gives better spatial coverage).
Second, the rainfall observations combine information from a variety of satellite instruments, including radars as well as instruments measuring microwave and infrared radiation.
Finally, ten years of surface wind measurements over tropical ocean regions were analysed to link large-scale circulations to rainfall and sea surface temperature. The wind product comes from a satellite scatterometer (opens in a new tab), which is a microwave radar that uses ocean roughness patterns to estimate the speed and direction of the surface wind.
For more information on the satellite contribution to wind and wave forecasts for offshore operations and ship routing, take a look at an ESA article: Wind and wave forecasts for offshore operations and ship routing (opens in a new tab).
Figure 1: ASCAT (a Scatterometer instrument (opens in a new tab) that looks at surface winds, along with GOES visible satellite information for 17 Sept 2020, Hurricane Teddy: at this point it was intensifying into a strong hurricane, and located about 610 Miles …980 KM east-northeast of the lesser Antilles islands of the Caribbean (e.g. Atigua and Barbuda). Source: CIMSS Satellite blog (opens in a new tab)
Good, P., Chadwick, R., Holloway, C.E., Kennedy, J., Lowe, J. A., Roehrig, R. and Rushley, S. S. High sensitivity of tropical precipitation to local sea-surface temperature. Nature (2020). doi: 10.1038/s41586-020-2887-3. Free read-only PDF (opens in a new tab).