The ECOSENSE forest – enriching tower-based flux measurements of carbon and water exchange with novel distributed sensor networks

ECOSENSE is a Collaborative Research Centre (CRC) funded by the German Research Foundation DFG to develop and test novel environmental sensing techniques and apply them to enhance our understanding and modelling of water and carbon fluxes in forest ecosystems. Specifically, with a distributed sampling approach, ECOSENSE studies the impact of forest heterogeneity in space and time in response to hydro-climatic stresses (Werner et al. 2024). Core to the project is the ECOSENSE forest, located in the foothills of the Black Forest in Germany (48.2685 N, 7.8782 E). The forest primarily consists of Beech ( Fagus sylvatica ) and Douglas-fir ( Pseudotsuga menziesii ). Two 30 m canopy access towers, one in a F. sylvatica and the other in a P. menziesii plot, and a 46 m tall eddy-covariance tower, which exceeds the canopy height by 20 m, form the core infrastructure. The integrated tower-based eddy covariance measurements are complemented with in-situ and remote sensing sensor networks distributed throughout the forest. The sensor networks provide simultaneous data on soil, litter, subcanopy, tree and leaf conditions. Cameras, drones and LiDAR systems record spectral and morphometric properties from seconds to seasons. In this contribution, we highlight how tower-based eddy-covariance flux measurements, which spatially integrated carbon and water fluxes above the mixed forest, benefit from massively distributed sensor networks to account for identifying the horizontally and vertically heterogeneous processes at the soil, tree and leaf-level. This will allow for the detection and attribution of “hot spots” and “hot moments”. We show the potential of integrating sensor networks and flux tower measurements, using data from the first year of continuous and parallel measurements at the ECOSENSE forest. 400 near-surface sensors and 20 vertical depth profiles continuously measure soil water content and soil temperature dynamics with unprecedented spatial resolution across the flux tower footprint. Sensors are deployed in different tree species plots and strategically measure gradients close to tree stems vs. further away. In addition at 13 sites, continuous soil carbon-dioxide concentrations are measured. Sub-canopy eddy covariance systems at 2 m height measure soil carbon and water fluxes in an area dominated by F. sylvatica and in an area dominated by P. menziesii . As an example of the potential of this dataset, we show how ecosystem-integrated carbon fluxes can be linked to the horizontal distribution of soil respiration, gross primary productivity (GPP), soil moisture and temperature, and assess the relative importance of soil water in controlling and dominating ecosystem fluxes under the same climatic forcing. At the tree level, a distributed sensor network consisting of 54 sap flow sensors (heat-pulse-method), 34 point dendrometers and 18 micro tensiometers provide tree-specific information on water fluxes that will be linked to tower-based evapotranspiration measurements. At the leaf-level, little branchlets of P. menziesii and leaves of F. sylvatica are equipped with novel, lightweight and non-invasive cuvettes (12), on six different trees while measuring temperature and humidity inside the cuvette, in ambient air, and on the leaf surface. The cuvettes are connected via PTFE-tubing to an automated air sampling system to continuously measure leaf CO 2 assimilation, 13 C discrimination and transpiration in the sun and shade canopy. Additionally, biogenic volatile organic compound (BVOC) emissions from the cuvettes and from nine vertical canopy layers are sampled campaign-wise and analysed via GC-MS. To measure continuous active chlorophyll fluorescence at the leaf-level, multiple MICRO-PAM measuring heads are attached in the sun canopy of P. menziesii (n = 3) and F. sylvatica (n = 3), directly accessible from the tower platforms. The measuring heads acquired all chlorophyll fluorescence parameters, leaf-level photosynthetically active radiation and leaf temperature at 15-minute intervals. We use this information to compare leaf- and ecosystem-level water use efficiencies and test different methods to partition ecosystem eddy fluxes into GPP. Also leaf-level BVOC fluxes are compared to ecosystem-level fluxes determined by a vertical gradient system and an experimental BVOC relaxed eddy accumulation system on the 46 m tower. Leaf-level active chlorophyll fluorescence sensors are compared against continuous tower-based spectroradiometric measurements (FLOX) to calculate canopy-integrated sun-induced fluorescence signals, which again are linked to eddy-covariance derived GPP. Distributed sensors are also installed to track leaf structural and leaf water content dynamics. Water-related leaf angle variations during day and night are recorded as image time series from 24 cameras and analysed using a deep learning approach (AngleCam). Additionally, six GNSS antennas, installed on the ground below the canopy and one above on the tower to track vegetation optical depth, provide high-temporal-resolution estimates on the variation of total leaf foliage area and canopy water content. We conduct monthly repetitive laser scanning as well as spectral imaging observations using unmanned aerial vehicles to gather high-resolution data on vegetation structure and vitality within the central eddy footprint area. This data allows us to continuously track specific tree properties such as tree heights and enables the analysis of seasonal changes and stress-induced variations, which is another link to longer-term tower flux measurement trends in the future. All data is collected on-site and transmitted in near-real time to a central database and visualization platforms. The data are used to initiate, drive, evaluate and further develop physiologically-based ecosystem models. In summary, combining integrative tower measurements with microscale sensor networks allows us to observe ecosystem processes across scales and time.