Thanks to AI: Forest growth determined from old data from Landsat satellites
Until now, it was not possible to determine how quickly and how high forests have grown from old satellite data. A research team has now changed that.
(Image: adamikarl/Shutterstock.com)
A research group has succeeded in using machine learning methods to determine the average canopy height in South Chinese forests for more than 30 years into the past, thereby contributing an important tool in the fight against climate change. The Chinese Academy of Sciences, which led the work, explains this now. For this, data from the US-American Landsat missions were evaluated, which made the time series possible in the first place. The analysis has thus revealed significant differences between tree growth in plantations and secondary forests and shown how important forest management is. The new method could be particularly helpful in this regard in the future.
Of High Value for Research
As the research group explains, maps of tree height exist for the world's forests. These are important for determining the biomass present there and finding out how much carbon dioxide the plants can store there. However, these maps always only represent the state at a specific moment, and insights into the dynamics in these important ecosystems could only be gathered to a limited extent. It is precisely these that are to be derived from historical satellite images, the team writes. This allows, for example, to determine the consequences of actions on site, both present and past. The study was presented in the journal Journal of Remote Sensing.
Videos by heise
Using forests in South China as an example, the research group demonstrated the functionality of its method and has now also published the results. According to this, the average canopy height there grew by more than 60 percent from 6.4 meters in 1986 to more than 10.3 meters in 2019. Large-scale reforestation projects and forest protection measures have thus ensured that areas with predominantly taller trees have spread significantly. Plantation forests grew significantly faster, but secondary forests ultimately reached greater overall heights. The team explains that the method can be applied worldwide and states that it can transform forest management.
(mho)
