Ecosystem phenology and productivity are described by GEO BON as the duration and magnitude of cyclic processes and the rate at which energy is transformed to organic matter at the ecosystem level. Changes in phenology and productivity can provide indications on important ecological changes (Schmeller et al. 2018) and on how ecosystems respond to climate change and disturbances, such as land degradation (Sims et al. 2019). Ecosystem phenology and productivity can be informed by remote sensing-enabled biodiversity variables (Skidmore et al. 2021). Here, we present case studies on how to utilize remotely sensed phenology and productivity for (i) the mapping of habitat types in northern Lapland and (ii) the modelling of forest pests at national scale.

Large parts of Finnish fell habitat types are endangered. Climate change, reindeer grazing pressure and pest outbreaks cause changes in habitats with high biodiversity value. To maintain and restore the conservation values of threatened habitats, accurate information on the distribution and status of habitat types are required. For large areas this information cannot be obtained by field surveys. We used remote sensing data and machine learning techniques to update the habitats databases in northern Lapland. Phenological statistical features from Sentinel-2 time series were used in addition to Sentinel 1,2 mosaics and laser scanning data in the prediction of habitat types. These features were found important in the prediction since they described well the changes in productivity levels in the inventory classes. For the prediction of forest pest outbreaks in Finland, we tested logistic regression modelling using green-up date from the Moderate Resolution Imaging Spectroradiometer (MODIS) as explanatory variable in addition to climate observations. According to our findings, green-up supported outbreak predictions for several moth species. It was particularly important for species that fly early in spring, and it can also play a role in larvae survival for species that hatch during the green-up period.  
Our result show that remotely sensed phenology and productivity can support the monitoring of changes in ecosystems and the modelling of species abundances for large areas. We will discuss limitations of the approach in Finnish conditions and future development and research directions.


Schmeller, et al., 2018. Biological Reviews 93, 55-71.
Sims, et al., 2019. Environmental Science & Policy 92, 349-355.
Skidmore, et al. 2021. Nature Ecology & Evolution 5, 896-906.