Estimating data, measuring impact
Corona is not over. The virus will continue to impinge on our lives until there is a vaccine available in sufficient quantity. In the meantime, Jülich scientists are researching how the restrictions of everyday life affect the environment, and they are improving predictions on the spread of the virus.
High season for climate research
Jülich climate researchers seized the opportunity: they launched extensive measurement campaigns to document the change in air quality during the lockdown in spring and also in the period afterwards. Equipped with Jülich instruments, two weather balloons rose to heights of up to 29 kilometres. A Zeppelin NT equipped with high-tech Jülich gear was used for the ground-level measurements between 200 and 1,000 metres above sea level, while the Jülich MobiLab measurement vehicle collected comparison samples on the ground. The campaigns recorded values for a large number of trace gases as well as carbon monoxide, nitrogen oxides, CO2, CFCs, methane, water vapour, ozone and the particulate matter content of the different air layers.
Initial results show that the lockdown had no major impact on CO2 levels in the upper troposphere and stratosphere at an altitude of around 15 kilometres. “We’ve been pumping billions of tons of carbon dioxide into the air for over 200 years – one cannot expect it to be significantly less or even gone in a few weeks,” says Dr. Johannes Laube from the Institute of Energy and Climate Research plainly. Further measurements and the evaluation of all data are currently underway. The results are incorporated into several climate models to close knowledge gaps and to predict, for example, the effects of a radical change in transport on air quality in different regions.
Corona predictions updated daily
How are the infection figures developing in my region or at the planned domestic destination? Answers to this question are provided by the results of a model made available by neuroinformatics scientists from Osnabrück University and data specialists from Forschungszentrum Jülich. Based on data from the Robert Koch Institute, the model calculates a five-day forecast for each district in Germany. The results can be viewed interactively on the Internet. The model also includes infection events from neighbouring regions in order to better assess the dynamics of the spread.
Access to the district-based corona forecast (in German):
more electricity: In March 2020, at the beginning of the corona crisis, photovoltaic systems in Delhi, India, achieved significantly higher yields than is usually the case at this time of year. Scientists from the Jülich branch of the Helmholtz Institute Erlangen- Nürnberg, together with partners from the Massachusetts Institute of Technology and the solar company Cleantech Solar, found that the decrease in man-made air pollution resulting from the corona lockdown allowed more sunlight to pierce through to the photovoltaic systems. The researchers had expected this, but were surprised at how considerable the effect was.
3 questions for …
... Laura Helleckes and Michael Osthege, doctoral researchers at the Institute of Bio- and Geosciences (IBG-1). The biotechnologists help to provide estimates of the daily R value for Germany and the USA. This value indicates how many people an infected person infects with the novel coronavirus on average in a certain timeframe.
How is it that biotechnologists are concerned with the R value?
Laura Helleckes: In our research, hundreds of automatically running experiments on the growth of bacteria have to be evaluated promptly and translated into models. We develop the necessary software tools for automating statistical evaluations. For this, we use an open source module called PyMC3. This is the same library that the creators of the website https://rt.live use for their R value estimates in the USA.
And you then came into contact with them?
Michael Osthege:Exactly. rt.live had technical difficulties and Instagram founder Kevin Systrom, one of the brains behind rt.live, approached the PyMC3 development team. We became aware of the project this way and began to contribute our expertise.
Aren’t growth experiments and the R value two different things?
Michael Osthege: Mathematically, there isn’t much difference in the evaluation of such data. Our experience in automating this type of analysis could therefore easily be transferred. It even turned out that we could learn a few things from the Rtlive model for the evaluation of our own experiments.
Laura Helleckes: We then applied the knowledge to German data and additionally developed the website https://rtlive.de with daily forecasts of the R value for German federal states.
You can find the complete interview here:
More about Jülich’s corona research at:
Texts: Brigitte Stahl-Busse
Photos: Forschungszentrum Jülich/Ralf-Uwe Limbach, Matteo Migliorati/Shutterstock.com, Lightspring/shutterstock.com