Application of the process based model 3D-CMCC-FEM to simulate the productivity at regional scale in Basilicata (Italy)

The Mediterranean basin is a well-known drought-prone region, nevertheless it has been affected by unprecedented extreme climate events over the last decades which are projected to lead to an even drier and warmer climate in this region. Process-based forest models are versatile tools being able to simulate forest stand growth and structure under a wide range of climatic and environmental conditions and can thus help to enhance our understanding of response to climate extremes over the long-term. Yet, such models are typically applied at a very local scale, ie. the stand level (or single sites) where high-quality data to initialize the model are available. Data scarcity at higher geographical scales represents a major challenge in upscaling such models, hence missing the opportunity to investigate the forest behavior over larger spatial extent and across multiple ecological gradients. For the first time, the process-based model 3D-CMCC-FEM was run on a regular 1x1 km grid over the Basilicata region, one of the Southernmost regions in Italy. The model was initialized using spatial information derived by the national forest inventory of the year 2005 and remote sensing data and regional forest maps. Carbon fluxes and pools have then been simulated over the period 2005-2019. In particular, we focused on gross primary productivity as the first proxy to potentially detect drought-induced damages on these temperate forests. The comparison to remote-sensing based data and observed growing volume stocks data available at selected sites, shows the capability of the model to simulate current forest productivity and to capture the forest response to summer droughts. These very promising results open the possibility to investigate the role of stand structure and species physiology on the spatio-temporal dynamics of forest growth under drought conditions over larger spatial scales.