It could be just the awe-inspiring images captured by Earth-observing satellites, which demonstrate undoubtedly the beauty of the Earth. But, there is so much more than that: what makes the Satellite Technology so valuable is its crucial role on gathering information about the complexity of the Earth system and offering a holistic view of the environment, resources, built –up areas. Armed with this data, the distinguished Professor Dr. Xiaoxiang Zhu, Head of Department “EO Data Science”, at the German Aerospace Center, aims to develop explorative algorithms, in order to answer important geoscientific questions, such as understanding the patterns of global urbanization. In her interview below, she refers –among other things- to her current research activities while she underlines the significance of Earth Observation for our future.
Ms. Professor Zhu, you work on signal processing for Earth Observation. What motivated you to focus on environmental data science?
Space has fascinated me since I was a child. I was also interested in mathematically demanding tasks from a very early age; as a schoolgirl I took part in mathematics competitions and always enjoyed dealing with challenging tasks. On the other hand, I would like to make a difference with my research, preferably in an interdisciplinary team. So what I am researching on today is no coincidence: The use of satellite technology, as it occurs in earth observation, perfectly combines my passion for space with my mathematical skills and interests. In addition, the working conditions at the German Aerospace Center (DLR) and the Technical University of Munich (TUM) are fantastic: As an Earth observation scientist, at DLR I am involved in large satellite projects and have access to up-to-date data from the most modern satellites, on the other hand, at TUM, I can train students, inspire them for my research topics and win them as doctoral candidates.
Can you kindly walk us to your current research activities?
I have been awarded an ERC Starting Grant, which will keep me scientifically busy for the next few years. In this project, methods and algorithms of machine learning are to be developed to fuse Peta bytes of remote sensing data from different satellite missions with massive image data and text files from social networks. This enables cities worldwide and their developments in 3D and 4D to be mapped, buildings/infrastructure types to be classified, population density to be estimated transparently, both in high resolution and globally. The result should be a new kind of global city model, which is, for example, particularly important for developing countries with their rapidly growing cities and slums.
A new kind of global city model, which is, particularly important for developing countries with their rapidly growing cities and slums, should be the result of the ERC Starting Grant, a project I have been awarded
On the other hand, Earth observation has irreversibly arrived in the Big Data era with the Sentinel satellites (and in the future with Tandem-L). This requires not only new technological approaches to manage large amounts of data, but also new analysis methods. Since April 2018, I have been promoted as the head of the department EO data science at the Earth Observation Center of DLR. Our mission is to develop explorative algorithms, in particular methods from artificial intelligence and data science, for the big Earth data analytics in general, aiming at answering important geoscientific research questions, for which understanding global urbanization is one example.
At what extent has our knowledge been advanced so far with respect to the behavior of the Earth system in comparison to the previous years?
Earth observation is satellite-driven, aims at digitalizing the dynamic Earth. Advanced sensor and mission concepts aim at higher resolution or/and larger spatial or temporal coverage. Both trends will eventually converge to very high resolution datasets on a global scale. Thanks to the European Copernicus program, nowadays it is possible to map the entire Earth with a spatial resolution of about 10m every week. As a prominent example of active sensors, the future German Tandem-L mission will possess a global coverage twice a week with a spatial resolution of 1 m. It will provide urgently needed information for solving pressing scientific questions in the areas of the biosphere, geosphere, cryosphere, and hydrosphere, and thus make a vital contribution towards a better understanding of the earth system and its dynamics.
Thanks to the European Copernicus program, nowadays it is possible to map the entire Earth with a spatial resolution of about 10m every week
And which in particular are the benefits for the society that the Earth remote sensing technology offers?
Geoinformation derived from Earth observation satellite data is indispensable for many scientific, governmental and planning tasks. Cartography, Geophysics, resource management, civil security, disaster relief, as well as for planning and decision support are just a few examples. As a concrete example, the global urban model I am deriving in the previously mentioned project is crucial to meet the information needs posted by the megatrend of global urbanization. For instance, although every third city dweller across the globe lives in slums we have no global knowledge on the dimensions, patterns, morphologies and locations of those. Even the population densities in slums are suspected to be grossly underestimated by current official statistics. Only with the information we are deriving it is possible to scale the fundamental infrastructure, such as health care, clean water and education, in these informal settlements according to the actual needs.
Cartography, Geophysics, resource management, civil security, disaster relief, as well as planning and decision support are just some of the benefits of the Earth remote sensing technology
Ms. Prof. Zhu, in which way do spatial data processing systems differentiate from the traditional ones? Which are the challenges that you and your team face in the arena of spatial data analysis?
As mentioned already, due to the advancement of sensor technologies, Earth observation has become a big data topic par excellence. Peta Bytes of data are freely accessible to every scientist. It becomes very challenging to dig out important information out from these data. Data mining and knowledge discovery, data analytics, high performance computing, as well as intelligent processing system are becoming indispensable components of our research. On the other hand, we — Earth observation scientists – are very excited about the new possibilities the big Earth data brought.
Based on your knowledge and experience on the Earth system and its complexities, which are the most important issues humankind is likely face within the next decades?
Climate change and global urbanization remain the most important megatrends. For example, according to UN, by 2050, around three quarters of the world’s population will live in cities. Although a strong correlation is verified between urbanization and prosperity of societies, urbanization does not automatically lead to a golden future of mankind. The new dimension of ongoing global migration into the cities poses fundamental challenges to our societies across the globe.
Although a strong correlation is verified between urbanization and prosperity of societies, urbanization does not automatically lead to a golden future of mankind
Could you share with us your vision about the future of Earth Observation?
Since decades, Earth observation is centralized and relies on satellite missions operated by space agencies, who provide geodetically accurate large scale geo-information worldwide on a routine basis from space. But the data availability is limited in resolution and viewing geometry. Recently, constellations of small and less expensive satellites owned by commercial players, like “Planet”, have been providing images for global coverage on a daily basis since 2017, yet with reduced geometric and radiometric accuracy. As complementary sources of geo-information massive imagery, text messages and GIS data from open platforms and social media form a temporally quasi-seamless, spatially multi-perspective stream, but with unknown and diverse quality. In the near future, all these developments will lead to community Earth observation: Earth observation becomes a field everybody can contribute by providing data sources, e.g., by uploading their street view images, every AI and data scientists can develop methods based on ready-to-use Earth observation data, and more people can directly benefit from it.