If global warming is not stopped, climate change will soon be noticeable on our plates. A third of the potato crop will be threatened by the end of the century - this is one result of Sabine Egerer’s research at the Helmholtz Center Hereon. She is trying to answer questions about the future of agriculture using data science methods - and has benefited from the support of the HIDA Trainee Network.
It is impossible to imagine German plates without the potato. Although consumption of the South American tuber is declining in Germany, according to the Federal Agency for Agriculture and Food (Bundesanstalt für Landwirtschaft und Ernährung - BLE), 55 kilos are still eaten per capita in this country every year. The demand for acreage is therefore still high. Climate scientist Sabine Egerer proves that the potato is not only of agricultural importance, but is also a true crossover research object. She has been working on the IMLAND research project at the Hamburg Climate Service Center (GERICS) of the Helmholtz Center Hereon since 2018, where her research on the potato combines questions of climate change and the future of agriculture with data science methods – to the benefit of Germany's most typical crop. .
Data science network for the next generation of scientists
Egerer is investigating the question of how climate change will affect potato yields in Germany and to what extent possible yield losses can be compensated for by irrigation. In order to calculate whether and how climate change will affect our plates in the future, extensive model calculations are necessary in which many variables must be included. However, since the Hamburg research group lacked expertise with statistical crop modeling, Egerer came up with the idea of getting help from elsewhere: She became aware of the Helmholtz Center for Environmental Research (UFZ) through a paper by scientist Michael Peichl, who models the influence of soil moisture on crop yields - also an interesting aspect for Egerer's potato problem. A colleague from Hamburg, in turn, pointed out to her the opportunity of networking professionally with colleagues at other Helmholtz centers via a guest stay and to get new ideas from the data sciences, with organizational and financial supported by the newly established HIDA Trainee Network.
This network, first launched in 2020, is aimed at PhD researchers and postdocs in the Helmholtz Association who want to expand their data science expertise across disciplines, and offers fully funded research fellowships for short-term stays at other Helmholtz centers. The goal of such exchanges: to benefit from each other's data science expertise, to network, and to learn about new approaches and ways of working. The home centers as well as the hosts can expand and strengthen their network within the Helmholtz Association through this collaboration. Despite the difficult Corona situation, which limited face-to-face exchanges, Egerer was highly motivated for a stay in Leipzig: "I really wanted to learn about new methods in machine learning and integrate soil moisture into my model. The Computational Hydrosystems group at UFZ are experts in hydrological modeling, and there's a lot of experience there in terms of data processing and machine learning algorithms."
Drought threatens potato crop
In north-eastern Lower Saxony, not far from the GERICS site, the problem Egerer is studying lies within reach: this is where the largest areas of potato cultivation in Germany are located. If a season turns out to be particularly dry, the potato harvest can suffer severe losses. Although this has only occurred occasionally so far, it is foreseeable that climate change will have a major impact on German potato cultivation in the future. This will have a direct economic repercussion on consumers, because prices will rise as supplies decline - and intensified irrigation of the fields will also incur costs and further deplete groundwater supplies. Since potatoes basically have to be irrigated artificially due to their shallow root system, irrigation is a key factor for future yields. But how drastically will climate change affect the crop and how could irrigation be adapted?
Modeling crop scenarios - in cooperation between Hamburg and Leipzig
For the Hamburg climate scientist, who was one of the first scientists to participate in the HIDA Trainee network between November 2020 and January 2021, the support of her colleagues from the Computational Hydrosystems Department at the UFZ in Leipzig came just in time. For her project entitled "Towards a statistical approach for agricultural yield production to evaluate climate change adaptation measures for present and future conditions", she developed a model in which the effect of irrigation on the harvest can be represented under different emission scenarios - with the help of the UFZ. In the process, she tested possible irrigation settings in order to determine the extent to which irrigation could compensate for future yield losses. Her colleagues also familiarized her with various data processing methods, Egerer said. "The greatest benefit of my exchange," the scientist is convinced, "was certainly the expansion of my knowledge of machine learning algorithms." Egerer was able to apply this new learning to select the most important climatic and hydrological variables responsible for crop yield variations. In addition, the scientist had access to much of the UFZ's data, such as climate data from the Helmholtz Climate Initiative, weather data from the German Weather Service, or data from the Soil Moisture Index (SMI), based on a model developed by the hosts. Regular virtual meetings enabled contact with the Leipzig team. Sabine Egerer is certain: "The official framework of the HIDA trainee network facilitated data access and scientific exchange."
The harvest of her stay at the UFZ: a successfully submitted abstract to the European Geophysical Union (EGU) and the presentation of her results at the EGU online conference. Together with the teams from Leipzig and Hamburg, a publication is planned in addition. "The most important insight for me is that a fruitful exchange is possible even without extensive direct contact", she says. Her verdict on her stay in Leipzig is unequivocal: "I would definitely recommend participation in the HIDA Trainee Network to other scientists, especially if they want to include external expertise in their work, find new collaborations and gain new ideas."
Providing stable yields: More water or fewer emissions
But what about the German potato actually, this complex research subject? Egerer's model shows that in the business-as-usual emissions scenario, which results in the highest temperature increase, high crop losses of over one-third are to be expected by the end of the 21st century without irrigation. With conventional irrigation, these losses are only reduced to 23%, or to 17% if the amount of water is doubled. A high price. So what the harvest will actually look like and what additional costs will be incurred in the future will depend very much on how much CO2 emissions can be curbed. Egerer's model, to which the UFZ-GERICS collaboration contributed, thus sharpens the view of the climate future in agriculture.
However, participation in the HIDA Trainee Network was fruitful for Sabine Egerer in another sense besides the successful research collaboration: "It became clear to me that I would like to continue working with statistical methods and machine learning algorithms in such a relevant field as agriculture. In this sense, the exchange has helped me to concretize my future plans in science."
Curious? Those interested in the HIDA Trainee Network can apply to participate again in the next round, which is expected to start in summer 2021. More info on the preparation and exact dates of the next call can be found here.
How can we tell at a relatively early stage whether a Covid-19 patient might develop a severe course? Data scientists tried to find an answer during the HIDA International Virtual COVID-Data Challenge.
Data challenges can help make the development of new algorithms more effective by improving the comparability of data. Lena Meier-Hein from DKFZ spoke at this year's Big-DATA.AI Summit about insights from challenges on biomedical image analysis.
Threatened potato crop: Sabine Egerer from the Helmholtz Center Hereon uses data sciences to explore what the future of agriculture might look like in times of climate change – with focus on the potato and with help from the HIDA Trainee Network.