The need and challenges of jointly simulating vegetation structure, diversity and ecosystem functioning: a tropical forest perspective

Despite a long history of vegetation modelling, robustly simulating vegetation dynamics remains a complex but highly needed task. There is a growing consensus that vegetation models need to better integrate forest structure, diversity and ecosystem functioning to tackle this research challenge. However, this has long been hindered by a coarse-grained representation of vegetation and subsequent difficulties to assimilate field data. Here I will present recent developments in an individual- and trait-based model of forest dynamics, TROLL 4.0. I will discuss the modelling choices we made to jointly simulate carbon and water fluxes, leaf phenology, as well as individual tree size and trait distribution, and evaluate them against a range of field-based and remotely-sensed data at two Amazonian sites. Based on this example, I will finally discuss several challenges that remain to fully bridge the gap between plant ecology, vegetation remote sensing and ecosystem modelling, and to improve our understanding and predictive ability of vegetation contribution to the Earth’s system.