EXPERIENCES FROM THE PROJECT COURSE IN GEOINFORMATICS

We reviewed optical remote sensing time series (TS) studies for monitoring vegetation productivity across ecosystems. Methods were categorized into trend analysis, land surface phenology, and assimilation into statistical or dynamic vegetation models. Due to progress in machine learning, TS processing methods will diversify, while modelling strategies will advance towards holistic processing. We propose integrating methods into a digital twin to improve the understanding of vegetation dynamics.

Long-term global LAI/FPAR products provide the fundamental dataset for accessing vegetation dynamics and studying climate change. This study develops a sensor-independent LAI/FPAR climate data record based on the integration of Terra-MODIS/Aqua-MODIS/VIIRS LAI/FPAR standard products and applies advanced gap-filling techniques. The SI LAI/FPAR CDR provides a valuable resource for researchers studying vegetation dynamics and their relationship to climate change in the 21st century.

Fungus-growing termites recycle large parts of dead plant material in African savannas and are significant sources of greenhouse gases. We measured CO₂ and CH₄ fluxes from their mounds and surrounding soils in open and closed habitats. The fluxes scale with mound volume. The results show that emissions from mounds of fungus-growing termites are more stable than those from other termites. The soil fluxes around the mound are affected by the termite colonies at up to 2 m distance from the mound.

We explore the rate of past and future sea level rise at the Finnish coast, northeastern Baltic Sea, in 1901–2100. For this analysis, we use tide gauge observations, modelling results, and a probabilistic method to combine information from several sea level rise projections. We provide projections of local mean sea level by 2100 as probability distributions. The results can be used in adaptation planning in various sectors with different risk tolerance, e.g. land use planning or nuclear safety.