Jens Wickert is head of the GNSS Remote Sensing Research Unit at GFZ department Geodesy, with his research focus being the GNSS radio occultation. GNSS stands for 'Global Navigation Satellite System' and is the generic term for navigation systems like the European Galileo or the American GPS. We take the launch of a new satellite mission, COSMIC 2, as an opportunity to talk to Jens Wickert about his research.
What is GNSS remote sensing?
We use radio signals from navigation satellites, originally intended for positioning, to determine geophysical properties of the atmosphere and the Earth's surface.
The GNSS signals are very accurate and do not have to be calibrated, as is the case with many other measurement methods. The signals change as they pass through the atmosphere. We analyze this change in order to derive, for example, the water vapor content or the temperature.
What do the individual working groups in your research area do: GNSS meteorology, radio occultation, and reflectometry?
Under GNSS meteorology we summarize all measurements with ground-based receivers. In Germany there is a network of about 300 stations, which we use in near-real time to derive the water vapour content in the atmosphere.
Behind this is the German satellite positioning service SAPOS. The data is made available to surveying offices. We use it in parallel for the determination of water vapour and make the resulting data available to the German Weather Service.
And what about GNSS reflectometry and radio occultation?
In GNSS reflectometry, the signals reflected by surfaces – such as soil, vegetation, water or ice - are analyzed to derive their properties. The roughness of a water surface can, e.g., be used to characterize wind conditions.
In GNSS radio occultation, measurements are made aboard satellites. They fly at around 400 to 600 kilometres above the Earth and receive radio signals from the navigation satellites, which are around 20,000 kilometres higher.
This is my personal scientific background. In 1999, I started my PhD as a physicist at GFZ with the analysis of CHAMP data, being one of the first in the world to use this method. CHAMP was a satellite mission, which was initiated and led by GFZ scientists.
Is the use of GNSS data part of weather forecasting?
Yes, radio occultation data from TerraSAR-X and TanDEM-X satellites, analyzed by GFZ, are used day by day in global forecasts of weather centers. They are among the most important and accurate observations. Also the radio occultation data from the American-German satellite mission GRACE-FO will be operationally provided.
The measurements of the SAPOS station network are used for regional forecasts, for example of precipitation. There is no other method for water vapor observation that can provide the necessary information about Germany in a better spatial and temporal resolution than GNSS with the SAPOS ground stations.
So far, the forecast is still rather inaccurate, especially with regard to heavy rain or locally limited events. This is noticeable as a user of weather forecast apps, where the temperature forecast is already quite good, but the rain forecast is not.
Are you working on an improvement?
In a current project, funded by the German Research Foundation, we plan to improve precipitation forecasting for Germany developing next generation GNSS based meteorological data products. For this we not only look vertically upwards, but also diagonally in different directions to all visible GNSS satellites at the same time. This gives us much more information about the atmosphere and we plan to make it available even faster than before.
There are rain radar apps for several years now. What about them?
If you want to know if it will rain tomorrow evening, these apps still quite often don’t deliver reliable results, yet. Behind this are forecasting models, which are not optimal and do not work with sufficiently accurate data. Our research results from GNSS meteorology contribute to better observational data and will therefore be able to improve rain forecast in the future.
How did measurements of GNSS signals by other satellites, i.e. GNSS radio occultation, begin?
In June 2000 the CHAMP satellite was launched. It was planned to be equipped with a GNSS receiver for positioning, and gravity measurements, anyway. This receiver was in addition appropriate to test the radio occultation method.
At that time there was no one in Germany able to process these measurements. Through project funding from Helmholtz for particularly promising new methods, I was hired as a scientist at GFZ, directly after working for the German Aerospace Center in the CHAMPproject, where I had started working on the method. I was in charge of the publication of CHAMP’s first radio occultation results.
At that time, we had a very ambitious and capable team here at GFZ and the data was analyzed faster than at NASA’s, which was also involved in the mission and whose scientists were far more experienced. We were also lucky the data was so good. Satellite missions are extremely complex undertakings, and a lot can go wrong. Imperfect hardware cannot be replaced once the mission is launched.
What does the data analysis look like?
We have developed software packages that carry out the calculations. On the one end the measurements of the radio signals come in, on the other end the temperature and water vapor values come out. We compared the resulting meteorological CHAMP data with measurements from weather balloons in order to assess their accuracy.
The evaluation runs automatically, not just during the day when someone is sitting in front of the computer, but permanently. Our software system is designed in such way that it regularly "checks" whether new data are available and then starts to process them automatically. That was also something new at the time.
What happened after CHAMP?
In 2006, the Taiwanese-American mission COSMIC was launched, with six small satellites receiving GNSS signals. It was the first satellite mission launched specifically for the purpose of GNSS radio occultation. I advised the colleagues there during the planning phase, starting in 2000. Our experience from the CHAMP mission was very important for this.
Other countries followed: There are now three such satellites of the European organisation EUMETSAT, but also countries like Brazil, China, Japan, India, and South Korea prepared and launched similar missions. I have supported several of these missions.
The CHAMP mission was the first of this kind and a real success story.
This year, the COSMIC-2 mission was launched.
That's right. Again with six satellites flying initially at 550 kilometers altitude. More than 4000 measurements are expected per day, which is enormous and much more than before.
The satellites provide mainly data from the equator and the mid-latitudes. The main goal is to improve typhoon forecasting. Taiwan, for example, is exposed to an average of four to five severe typhoons per year. Also for Space Weather related investigations the equator region is very important.
Compared to COSMIC-1, the satellites are larger, the receivers are much better, and the signal strength of the data, sent by the COSMIC-2 satellites to Earth is also much higher.
To what extend were you involved here?
I was part of review panels, especially for COSMIC-1. A satellite mission involves huge sums of hundreds of millions of dollars. The money is made available step by step, when a planning or implementation phase has been successfully completed. The review panels are set up to assess this.
As the mission progresses, I participate in conferences where results are presented or evaluate research projects, where data from COSMIC-2 are used. Of course we also will work at GFZ with data from COSMIC-2 in our GNSS radio occultation group.
For 2023, Taiwan plans to launch a satellite for GNSS reflectometry, where we were asked for advice, again.
Why a satellite for GNSS reflectometry?
On a mission like CHAMP, the satellite only "looks" upwards or sideways through the atmosphere between itself and the navigation satellites. The reflectometry satellite receives the radio signals, reflected from the Earth's surface and can thus be used to study various surfaces such as water, land or ice.
Such satellites are not dependent on ground receivers and can collect data globally from space. Otherwise, it would be very difficult to obtain comprehensive data on parameters like wind or rain, especially over the oceans.
More and more satellites are also a danger: current buzzwort 'space debris'. Do we need all these new missions?
There are still observational gaps. And for weather and climate models, the data should become even more accurate in order to get better results.
Especially for the COSMIC-1 mission, the satellites are now simply "dead" after more than ten years of operation and are no longer transmitting data. This is due to factors such as the influence of radiation at high altitudes, which destroys the technical systems.
And especially for climate monitoring it is important to not have large gaps in the recording, but to measure continuously.
Where is GNSS remote sensing heading to in the future?
The future belongs to small and very small satellites. A large number of satellites can be launched into space at the same time, at very low costs. And if a satellite fails, it is much less tragic than on a mission with only one large satellite, where then the entire mission fails.
We are currently planning a small satellite project with the European Space Agency: PRETTY. This mission is based on a so called Cubesat, which consists of three segments, each 10 times 10 times 10 centimetres in size. It will orbit the Earth from 2022 onwards. PRETTY will collect GNSS reflectometry and space weather data.
However, the large weather satellites will retain their justification. They provide other data that we cannot get with GNSS, like cloud images.
Interview: Ariane Kujau