Satellite remote sensing presents an inevitable global source of synoptic data that enable retrospective characterization into the dynamics of the vegetated land surface and associated biogeophysical and biogeochemical processes. Remote sensing enables acquisition both of long-term observations and information about the intra-seasonal dynamics of vegetation, or “land surface phenology”. Since there are no existing nationwide studies with very high spatial sensitivity of the land cover changes as a consequence of historical political and socio-economic trends, this project project will use recent advances in state-of-the-art predictive modelling and machine learning. In that sense, the main scientific research objectives of the project are:
- To develop technical framework, based on the long-term Earth observation and advanced data processing, statistical and machine learning algorithms, for evaluation of the spatio-temporal land cover dynamics due to institutional changes and regional impacts of global environmental change.
- To provide results that enable precise evaluation of the impact, vulnerability and adaptation capacity of the main vegetation types and land-use classes (forests, semi natural areas and agriculture areas) across the regional environmental settings in Croatia.
- To develop early warning systems for agriculture and forestry based on near-real-time meteorological data and short-term forecasting of phenological responses.
The project is based on stated hypotheses:
- Is it possible to assess the long-term climatic and anthropogenic effects on the spatio-temporal vegetated land surface earth observation data?
- Is it possible to determine degree of adaptation of each vegetation type to climate change? Or simplified. Which forest type and agricultural crops are least and most affected by climate change?
- Is it possible to provide short-term forecasting of climate impacts on vegetation classes?