Earth observation is an area of the space sector with many applications. Access to regularly updated basic imagery is very useful – how many times have you used Google Maps in the last year? In a more complex way, though, Earth observation imagery can be analyzed to provide insights about what is happening on our planet. The potential phenomena which can be analyzed vary greatly, ranging from farming to armed conflict, for example. To learn more about Earth imagery analysis, we spoke to Henrik Fisser, a master’s student studying applied Earth observation at the Julius Maximilians University of Würzburg.
Why did you decide to do a master’s degree in applied Earth observation?
It was in Uganda when doing an internship in a remote rural region in the country’s northeast when I learned about the potential satellite data has for monitoring Earth. People living in that semi-arid Ugandan region are mostly farmers relying on rain. In 2017, when I was there, East Africa was hit by a drought. I recognized that information derived from Earth observation data might support prediction and management of such disasters. In the final thesis of my geography undergraduate, I thus analyzed how the land cover of the region had changed since 2000. It turned out that more soil was being exposed – a sign of desertification. I was sure that analyzing such data is what I want to specialize in. Deciding to pursue my master’s degree in applied Earth observation has definitely been the right choice.
What have you learned so far about Earth observation that has surprised you?
At the beginning of my studies, I was surprised by how hard Earth observation data science is. There is much to consider when analyzing this data. Already during the final thesis of my undergraduate, I had read about atmospheric correction, radiometric calibration, cloud masking, and other data preprocessing steps. It surprised me at the beginning how much effort it requires to derive meaningful, valid information.
However, I also learned that people around the globe are working on reducing the gap between data providers and users. This decreases the preprocessing overhead and gives scientists and other users more time and resources to do their actual analyses. Actually, some years ago, much of the data was either nonexistent or expensive. Earth observation data is much more accessible than it was ten years ago.
I personally got interested in making data screening and selection more efficient. Thus, I developed an algorithm that computes the cloud cover of Earth observation data in a user-provided area without downloading large amounts of data. It just works on small quick-look images. I hope that this work that I do together with a fellow student will help people, once we make it available as a software package.
What is one of the most interesting things on Earth you have observed during your studies?
For a term paper, I analyzed how we can monitor the impact of war from space. Yemen experiences a horrible proxy war. I was curious if and how we can monitor some dynamics of the war in the Yemeni port of Hudeidah, which is critical for humanitarian goods entering the country. In fact, using Sentinel-2 data of the European Space Agency, I could demonstrate how the port was hosting less activity after rebels had seized it. It was fascinating to be able to monitor this from my safe desk in Europe.
That was just for a term paper, but such observations have the potential to monitor conflicts and ultimately help people. I want to develop and use my skills for issues that matter to people and the planet. Our home planet seen from space, be it described by an astronaut or captured by a satellite, it reminds us of how precious life is. You don’t have to be an Earth observation analyst to understand that.