You have been one of the most active users of Fluxnet data and using it to produce data based gridded data products using machine learning methods. What inspired you to go down this track?
We were discussing a lot within Markus Reichstein’s group on linking FLUXNET with remote sensing and a lot of inspiration comes from there. But it also happened a bit by chance. When I was working on interpreting FLUXNET data with a machine learning method I realized that these methods are powerful in producing good estimates while interpreting what’s going on inside the tool is quite challenging. … so I was a bit opportunistic here too.
What are some of the strengths, problems and limitations of developing machine learning data products?
I’m spending quite some time on figuring out what these products are good for and what not. In the beginning many people were quite skeptical about these products. Nowadays I have sometimes the impression that the products are taken as ‘truth‘, which is not useful. The strength of the products is that they provide a complementary global data stream derived from observations. Understanding of patterns and processes should ideally be based on the synergy of multiple data streams since each has its own limitations.
How has Fluxnet helped you achieve your goal and what other types of data and information from us do you need to do a better job into the future?
Almost all I did would not have been possible without FLUXNET. Without FLUXNET I might even be unemployed by now . All the ongoing efforts of expanding the network, harmonization, standardization, and uncertainty characterization along with the collection of meta-data are really important and will pay off soon.
What has been your path as a young person to being a scientist in our field?
There was a lot of chance and luck involved I have to admit. I was somehow always interested in and fascinated by environmental sciences and this together with the attitude of a little curious nerdy kid probably made me stay in academia.
Finally what words of advice do you have for junior scientists and from your perch what are the priorities we as a field should be focusing on and collaborating on?
Sharing data is very important for the field to gain big picture insights and patterns. There are many important and open questions such as adaptation and acclimation, CO2 fertilization and interactions with nutrients. I’m personally very interested in how carbon and water cycles interact and I’m curious about which role groundwater could play here.