Du
20 March
au
31 December 2013

L’imagerie hyperspectrale pour le suivi des prairies et le contrôle des mesures agro-environnementales

Hyperspectral imagery for monitoring grassland and for controlling agri-environmental measures

Context

The CASI-SASI 2002-2003 and CASI-ATM 2003 flight campaigns are proposed within the framework of the exploitation of instruments aboard aircraft within the Belgian STEREO program (OSTC) and the support of the European PRODEX program (ESA) for the development and the construction of a hyperspectral imaging spectrometer APEX (www.apex-esa.org). In this framework, the CRA-W participates in these three fligth campaigns to study the potentialities of hyperspectral imagers to assess grassland characteristics and to allow a monitoring at regional or local scales which is not possible with laboratory analysises. In contrast to multispectral instruments, hyperspectral imagers measure a great number of contiguous wavelength bands and could be a new remote sensing information tool for the agricultutral sector.

Objectives

With the CAP reform, the European agriculture should become competitive and at the same time environmentally friendly. Agri-environmental measures are now compulsory and consolidated in all rural development plans in the EU. Due to the diversity of farming systems typologies and their importance in the agricultural landscape, grassland and head-land concern an important part of the agri-environmental measures. Within the framework of a sustainable agriculture, monitoring of permanent meadows is closely linked to the knowledge of grasslands management system combined with their physico-chemical and environmental characteristics at the regional level. This information is generally collected by systematic ground observation campaigns and by laboratory analyses. Unfortunately, this approach needs a long time to be realized and can be expensive. A solution could be the use of remote sensing as consistent and non-intrusive verification tool in the management and the control of parcel-based agricultural activities. Nowadays, remote sensing is mainly used to control area and land-use (IACS systems), but due to their spectral resolution, the actual panchromatic or multispectral sensors are limited for ecological studies and for monitoring semi-natural grasslands. The spatial and spectral resolution of hyper-spectral sensors would significantly improve remote sensing capacities. The objective of this study is to show that hyperspectral imagery allows answering some management needs by supplying a continuous spatial and temporal monitoring of parameters that characterize the canopy structure of every grassland parcel as well as its biochemical or biophysical properties.

Results obtained

The first measurement campaign allowed to estimate the quantity (e.g. wet matter, biomass, grass height) and the quality (e.g. protein, VEM, DVE) of grass canopy and so to establish regional inventories on grass production capability. In addition to this, and based on textural analysis at the intra-parcel level, spectral response curves highlight a potential discrimination and classification between the various types of meadows (pasture, mowed meadows, etc.). All of this derivate information constitutes an important decision tool for farmers and local agencies. Using hyperspectrales time series imagery, the project now studies the possibilities of combining indices to find the relationship between the status of biophysical, the biochemical characteristics of grass and the time since the last cutting or start grazing. In an exploitation phase, these relationships will be used to estimate unknown dates of cutting or grazing with observed remote sensing data to provide a control tool for the application of the temporal constraints for the grass and headland agri-environmental measures. These first results are very promising and contribute to demonstrate new perspectives for use of Earth Observation in relation to up-to-date European Union requirements and the CAP evolution.

Contribution

*

Partners

Vlaamse Instelling voor Technologisch Onderzoek (VITO)

Funding

  • Federal Scientific Policy
  • ESA - European Space Agency