One of the scientific aspects to be further developed in the Forest Carbon Monitoring project is the integration of process-based forest ecosystem models into the forest biomass and carbon flux monitoring framework. The recent article ‘Demonstration of large area forest volume and primary production estimation approach based on Sentinel-2 imagery and process based ecosystem modelling’, published in the International Journal of Remote Sensing, gives a good overview of the concept to be integrated into the forest carbon platform. The study reported in the article was conducted in the earlier ESA Assesscarbon project, but the work will be continued in our current Forest Carbon Monitoring project.
The article demonstrates the functionality of a cloud based approach utilizing Sentinel-2 composite imagery and process based ecosystem model to produce large area forest volume and primary production estimates. The main components of the approach and implementation of the processing pipeline into the Forestry TEP cloud processing platform are described and four large area output maps are presented: 1) Growing stock volume (GSV), 2) Gross primary productivity (GPP), 3) Net primary productivity (NPP) and 4) Stem volume increment (SVI), covering Finland and the Russian boreal forests until the Ural Mountains in 10 m spatial resolution.
Since the forest structure information serves as the input for the process-based models, the estimation accuracy of the forest structural variables is of key importance for the system. In the Assesscarbon project, relative Root Mean Square Error (RMSE) values of 34-39% were reached for the key input variables (basal area, diameter and height) in pixel level. This is on a comparable level to earlier studies, although most of the earlier studies have concentrated on smaller study areas, while large area composite imagery was used in the reported study. This can be considered a positive sign for the feasibility of the approach for large area primary production modelling. The full coverage output maps showed consistent quality throughout the target area, with major regional variations clearly visible, and with noticeable fine details when zoomed into full resolution.
The demonstration reported in the article lays a good foundation for further development in the Forest Carbon Monitoring project. The improved estimation of forest structural variable is among the key scientific goals of the project. Possibilities of a combined use of optical and radar datasets will be evaluated and different types of algorithms (some of which are variable-specific, i.e. used only for estimation of a single variable like tree height) will be tested. With regards to the process-based models, the current project enables further calibration of the models for the conditions prevailing in different parts of Europe.
If you are interested in reading further about the process-based ecosystem models and their use in forest carbon flux estimation, you can also read a recent blog post ‘Are carbon flux products reliable?’ on the EU Forest Flux project website.
Full citation of the article: Miettinen, J., Carlier, S., Häme, L., Mäkelä, A., Minunno, F., Penttilä, J., Pisl, J., Rasinmäki, J., Rauste, Y., Seitsonen, L, Tian X. and Häme, T. (2021) Demonstration of large area forest volume and primary production estimation approach based on Sentinel-2 imagery and process based ecosystem modelling. International Journal of Remote Sensing 24: 9492-9514. doi: 10.1080/01431161.2021.1998715
