Process-based models are important tools to analyse the direct influence of climate and other environmental factors (e.g. soil properties, geochemical reactions, trophic relationships, species traits, CO2 concentration, land cover change, biotic and abiotic disturbances) on biological systems. Such models are built on mechanistic rules and allow to consider non-stationarity and non-linearity in ecosystems responses to specific environmental variables as well as their interactions. They can also be used to support management decisions and illustrate the impact of alternative management choices to stakeholders and citizens. Once the models are built, their performance can be refined by constraining them with data using model-data fusion techniques, i.e. calibration, assimilation, inversion. In this session, talks and posters describing model developments, comparisons with observed data, and progresses achieved in process understanding in all field of terrestrial biogeosciences are welcome. Different spatial (from the bacterial community to the landscape) and temporal (from the nanosecond to the millennia) scales can be analyzed.