Digital Twins
How deep learning and remote sensing can combine to promote 3D city modelling for digital twins
5 min
Since 2018, the Netherlands and many other parts of Europe have been faced with increasingly long drought periods. To understand how ecosystems respond to such climate extremes, soil and plants need to be studied in a coherent way as one system that reacts to these events. ITC researchers surmise that a digital twin of the soil-plant system would be an ideal means of observing and understanding everything that’s going on.
Digital twins are a technology that can help to push boundaries in developing a truly holistic approach to understanding processes. Digital twins are highly accurate virtual replicas of physical systems. They make it possible to simulate the system in a comprehensive manner and understand potential outcomes from possible scenario changes. A general concept stemming from industry, digital twins are used for all kinds of purposes. In the automotive and aircraft industries, for example, they eliminate the need for actual cars or airplanes to be damaged or destroyed in stress tests.
The health of the soil-plant system can't be judged by its appearance alone. There are processes at cellular scales to be taken into account: photosynthesis, stomata opening/closing, solar-induced fluorescence, but also radiative transfer processes at both the leaf and canopy levels. To understand what exactly is going on, it's not enough to look at the plant and the canopy alone. The soil is also important, because that's where water is uptaken by the roots and then supplied to the plant for evapotranspiration. Other crucial factors include soil-root interactions, soil matric potential, leaf water potential, and vapor pressure deficit, as well as the amount of nutrients present in the soil. It’s very important to consider all these factors and processes above and below the ground coherently. That is what the soil-plant digital twin is being developed to do.
The soil-plant digital twin can help to understand the impact of past events by simulating those events. It can also provide more insight into what is actually happening to the soil-plant system. What is the impact of prolonged droughts or extreme heat events? How can we predict and prevent certain unwanted effects in such a way that appropriate measures can be taken?
For instance, the soil-plant digital twin can be fed with weather-forecast data from a numerical weather prediction model. If the weather forecast says that a severe drought event will happen in two weeks, the soil-plant digital twin can help to predict the impact this event will have on plant productivity. This information can then be used for developing adaptation or mitigation measures to make the soil-plant system more resilient to climate extremes. As such, the soil-plant digital twin can help farmers to mitigate and adapt the impact of climate change on their livelihood, and show them how to improve their water and soil management to improve production in a sustainable way.
The digital twin now being developed by the Department of Water Resources at ITC is meant as a digital infrastructure for running two models that have been incorporated into one: soil model STEMMUS and vegetation model SCOPE, both (co-)developed by ITC researchers. The STEMMUS-SCOPE model aims to create a virtual continuum of the soil-plant system. Running the model in a traditional way would take much time and computing power, especially when upscaling from point to regional, continental or even global scale. The developed digital twin could overcome this problem via deploying the FAIR[1]-enabling digital technology, translating research needs and developments into reproducible and reusable software, data and knowledge. This soil-plant digital twin will facilitate an interactive and configurable platform that allows users to create and evaluate ‘What-if’ scenarios.
Want to know more? Check out our video on Soil-plant digital twins and how to model them.
[1] FAIR: Findable, Accessible, Interoperable, and Reusable