C6 Development of area-wide functional indicators using remotely sensed data.
PI(s) for this project:
Prof. Dr. Jörg Bendix
The main aim of this project is at the development and implementation of area-wide functional indicators (Evapotranspiration, Primary Production, and Water Use Efficiency) which can be used to monitor changes in tree water relations due to environmental change (land use and climate change) and to identify water sensitive indicator trees. Calibration and quality assessment of the algorithms will be conducted, among others, with scintillometry above the tree canopy and observation-based model results of project C5 and others. The two-big leaf approach will be considered, which requires tree crown illumination (sunlit and shaded fractions) to be calculated by means of LiDAR data. The second aim of this project is at the retrieval of area-wide structural and multi-/hyperspectral predictor variables from remote sensing data, which will be used to develop and implement area-wide functional indicators on biodiversity and ecosystem processes in collaboration with project C2. The third aim of this project is at the projection of future land use change scenarios based on change detection and Multi Objective Land Allocation (MOLA) techniques. The implementation in an operational monitoring system together with the non-university cooperation partners will be derived from operationally available data sources. Consequently, indicators and predictor variables will be developed on the crown scale with high resolution data (1 – 6.5 m resolution per Pixel) and then upscaled to the landscape scale (~30 m pixel resolution) for area-wide monitoring in southern Ecuador. Airborne and ground surveys will be (and have already been) conducted by the project, as e.g. Airborne Laser Scanning (ALS) and Hyperspectral Scanning, to support the development and validation of the area-wide functional indicators.
The main study site is located in the Reserva Biológica San Francisco (RBSF), on northern slopes of the valley of the San Francisco River, which are covered by the typical tropical mountain forest of the Andes of South Ecuador. In addition to the RBSF study site, two different ecosystems will be investigated in landscape and regional levels: The dry (seasonal) forest at Laipuna Reserve and the páramo highland vegetation at Cajas National Park.
WP1 Functional indicators
We will investigate heat fluxes, evapotranspiration, and productivity at the canopy scale on the core plot at the RBSF. Two observation towers (30 m height) will be erected in the forest core plot in the RBSF for measurements of heat fluxes and evapotranspiration (ET) at the leaf and canopy level. Automatic scintillometer and weather sensors will be installed on the top of the towers and on ground for measuring heat fluxes and evapotranspiration at the canopy level. Porometry at leaf level will be used for direct comparison with the canopy fluxes, including leaf photosynthesis. Individual tree crowns will be delineated to determine the indicators at crown level from high-resolution satellite data. Ancillary spectroscopy data will be carried out from the towers to obtain canopy spectral signatures and indices (e.g. water and vegetation indices), which will be used to calibrate/validate satellite data and to upscale the indicators from leaf to crown and landscape scales. Satellites (e.g. Worldview, Quickbird) will be deployed to acquire new high-resolution data throughout three years of monitoring. To cover the regional climate gradient one eddy covariance tower will be installed in Laipuna and Cajas each. Field campaigns will allow the characterization of the vegetation layer and the estimation of functional indicators will be pursued at landscape to regional level.
The WP2 aims at the determination of remote sensing data which can be used to derive area-wide indicators related to the biotic and functional indicator variables. The WP2 will focus the RBSF study area. We will provide variables related to nutrient/water stress and productivity, canopy structure (e.g. vegetation height and canopy density). Structural variables indicating vertical and horizontal canopy density will be derived from ALS data acquired with 10 returns per square meter. Spectral indices related to nutrient/water and physiological states and to biotic stressors (e.g. herbivory) will be derived with satellite and airborne spectral scanning. Together with field data (project C2) from the 90 x 90 m plots (see plot concept), we will develop transfer functions relating the status of biodiversity and biotic processes. Machine learning and partial least squares regression will be applied to calculate transfer functions, which will then be used to derive area-wide indicators using operational satellite data.
WP3 Land Use and Cover Change scenarios
The calculation of land use change projections is a service to C7 and will be conducted for Laipuna and the Cajas areas based on time-series of 30 m resolution satellite data (e.g. Landsat). Main indicators are cover changes for forest, agriculture and urban areas, which will be projected using Business as Usual scenarios in line with climate change prediction models.