Project Details
Making the direct link between light regime and forest biodiversity – a 3D spatially explicit modelling approach based on TLS, CNNs and ray tracing.
Applicant
Dr. Julian Frey
Subject Area
Forestry
Term
since 2020
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 448505589
Forests are ecosystems that are fundamentally defined by their three-dimensional structures. Important physical processes such as the availability of water and light and the temperatures in forest stands are mainly controlled by the structure and shape of the canopy. In turn, numerous ecosystem processes and services such as biomass growth, inter-specific competition or resistance to pests depend on these physical processes. Nevertheless, only incomplete understanding of how exactly and to what extent the structures of the forest influence these processes is available. For example, the technical hurdles to map the transmittance of light through the forest strata not only at specific points, but in their entire temporal and spatial dynamics are enormous. Models that address this problem have not yet been able to achieve a reliable small-scale representation of light availability in real stands. This is due to the fact that the three-dimensional representation of the forest was only insufficiently possible and that the subdivision of the biomass according to its transmission and reflection properties (foliage, branches, trunk, tree species) was not yet possible. Modern terrestrial laser scanning systems (TLS) currently achieve the precision, reliability and speed that make detailed recording of entire stands justifiable, and convolutional neural networks (CNNs) solve classification problems of new magnitudes. The goal of this project is to combine these two technologies to create sufficiently exact structural data a ray tracing based light model that can describe the distribution and quality of light in the stand in time and space in order to investigate processes based on this. In a further step we want to determine the distribution of surface temperatures from the radiation budget. For this purpose, TLS scans will be made on 20 investigation areas of 1ha each in southwest Germany in low mountain range forests and reference measurements with PAR sensors and thermal images will be carried out. In particular, the effects of different tree species mixes on the utilization of available light in the lower forest strata will be classified. In addition, we use taxonomic inventories (vascular plants, mosses and lichens, insects) carried out on the same plots by project partners to prove and quantify the direct influence of physical processes on the species abundance and distribution. Thus, parts of the species' ecological niches will be described on the same spatial and temporal scales on which the species occur. The goal is to develop a functional understanding of the connection between basic physical and ecological processes.
DFG Programme
Research Grants