Project

Regional ecosystem mapping

The Kyoto Protocol proposes a global policy for assessing carbon emissions and sequestration at an international level. The terrestrial environment is acknowledged to be one of the most critical components contributing to the dynamic changes in the global carbon cycle. Earth observation from airborne and spaceborne platforms is the most promising observational approach capable of providing data at the relevant scales and resolution needed to extrapolate findings of in situ (field) studies to larger areas.

mapping

The heterogeneity of the landscape at a regional scale can be characterised by statistical and/or GIS techniques, as well as by process modelling of ecosystems. The latter approach usually requires the input of spatially contiguous biophysical, biochemical and/or structural information on the ecosystem under consideration. Recent developments in the field of imaging spectroscopy enable mapping of biogeophyscial variables of the earth’s surface with unprecedented accuracy compared with traditional remote sensing techniques. Imaging spectrometers (sometimes referred to as hyperspectral sensors) are instruments that acquire images in many, very narrow contiguous bands so that for each pixel in the image it is possible to derive a complete refl ectance spectrum. The laboratory-like spectra from the ‘image cube’ can be used to quantify land cover properties. Recent advances in applying ecosystem process models prove that their combination with regional scale remote sensing is a very promising approach for testing ecological hypotheses and for assessing and forecasting the state of large landscapes.

Current developments

Recent advances in land biosphere modelling require not only Land Use/Cover Changes mapped to a high degree of accuracy, but also more sophisticated parameters derived by using advanced methods such as imaging spectrometry. In particular, the proper estimation of structural and biochemical parameters are of utmost importance. These currently include the determination of Plant Functional Types, chlorophyll a/b content and other leaf pigments, the C:N ratio in natural vegetated areas, and Leaf Area Index (LAI). The variables obtained from imaging spectroscopy can then be used to derive relevant input parameters (LUE, WUE, FAPAR, gap fraction, etc.) for ecological and climate modelling.

Mapping the often complex landscapes at a local or regional level usually requires airborne instruments which have the flexibility to map the full extent of such ecosystems (e.g., river floodplains) and ecotones (transitional zones) without too much data overhead. Spatial scaling to national or European applications is supported through the simultaneous acquisition of satellite data (such as MERIS on ENVISAT or MODIS on Terra/Aqua). On this scale, imaging spectroscopy can provide the long-term observations of the state and mechanisms of the biosphere required for European programmes such as GMES (Global Monitoring Monitoring of Environment and Security).

Application

A combined remote sensing/ecological modelling approach for a river floodplain in the Netherlands has been developed in a joint Belgian-Dutch project (VITO, VUB, ULB, Wageningen UR). Airborne imaging spectroscopy data from the HyMap sensor (HyVista Corp., Australia), acquired in the summer of 2004, were used to derive biophysical and biochemical products (e.g. Leaf Area Index, FAPAR, hydrocarbon index). The variables derived by imaging spectroscopy were used in the ecological model SMART-SUMO (developed within Wageningen UR) to simulate vegetation development under scenarios of changing abiotic conditions and management (e.g. grazing density). The remote sensing data were validated by destructive biomass sampling in 21 plots, in combination with VALERI sampling in the forests and canopy reflectance measurements. Different ‘vegetation’ scenarios were implemented to simulate total biomass in the floodplain in the year 2050. Significant simulation differences in total biomass arising from different management approaches (agriculture versus nature reserve) demonstrate the need to develop ‘vegetation’ scenarios analogous to IPCC atmospheric scenarios.

A crucial requirement for the advancement of imaging spectroscopy is having regular access to a platform offering support for these kinds of instruments, including the availability of field equipment (field spectrometers, GPS, etc.) to support ground data acquisition for validation purposes. The Centre for Geo-information, part of the Environmental Sciences Group of Wageningen University and Research Centre participates and contributes to major international developments in remote sensing. Signifi cant input to the further advancement of spectrodirectional imaging has been made through contributions to the ESA Earth Explorer Core Mission definition (SPECTRA), and to new sensor development (ESA APEX), defi nition (ESA SPECTRA, FLEX, Sentinel) and calibration (ESA MERIS/ENVISAT and CHRIS/PROBA). In addition, the Environmental Sciences Group has a field equipment pool with state-of-the-art instruments.

Future perspectives

The emerging applications of imaging spectroscopy of vegetation will focus on monitoring:

  • transitional zones, in particular ecotones (ecosystem or habitat boundaries, e.g. tundra–boreal forest), where most of the pressures and ecosystem disturbance are being identified;
  • managed ecosystems, in agricultural applications such as site-specific management where precision appliance is a key economic factor resulting in higher yields and reduced emissions to the environment;
  • managed or unmanaged ecosystems, where plant succession, plant functional types and invasive species are important focus areas.

The combination of regional scale remote sensing and biogeochemistry based process models to answer ecological and CO2 related questions will be a primary focus over the next few years. The major challenges in imaging spectroscopy are the mismatch of spatiotemporal scales of field, airborne and spaceborne measurements, and model requirements. The multidisciplinary expertise of the Environmental Sciences Group and its international connections provide an excellent framework for resolving these issues.