Evaluating the potential of crop specific LAI time series retrieved from various SAR sensors to support agricultural monitoring at regional scale based on optical imagery

Bériaux, Emilie;Defourny, Pierre;Duveiller Bogdan, Grégory
(2009) International GEO Workshop on Synthetic Aperture Radar (SAR) to Support Agricultural Monitoring — Location: Kananaskis, Alberta, Canada (2.November.2009)

Files

No attached file found for this publication.

Details

Authors
  • Bériaux, EmilieUCLouvain
    Author
  • Defourny, PierreUCLouvain
    Author
  • Duveiller Bogdan, GrégoryUCLouvain
    Author
Abstract
Green leaf area index (LAI) characterizes both the structure and the photosynthetic capacity of a crop canopy. This variable is therefore of great interest for agricultural monitoring since it allows the coupling of earth observation with crop growth modelling. LAI can effectively be estimated at different scales from optical remote sensing and several sensors provide a global coverage with high acquisition frequency providing a potential for regional agricultural monitoring. However, relying exclusively of optical data is hazardous due to the important cloud cover present in many places around the world during the crop growing season. This research provides insight on how information from different SAR sensors can assist and consolidate crop specific LAI products from optical medium resolution sensors when data availability is jeopardized by cloud cover. This research is realized in the framework of the international GLOBAM project, a globally distributed agricultural experiment to enhance crop monitoring by remote sensing. The study area is located in central Belgium where an intensive field campaign was synchronized with remote sensing acquisitions between April to July 2009. Ground LAI was regularly measured using hemispherical photography over a sample set of large fields for both maize and winter wheat. By interpolating temporally these LAI measures using a Canopy Structural Dynamic Model (CSDM), a ground truth time series for each field is constructed which serves as a benchmark for the LAI obtained from the various remote sensing instruments. LAI time series were retrieved from different types of both SAR (ERS/SAR, ENVISAT/ASAR, RADARSAT-2) and optical (SPOT4, MODIS) imagery. LAI is retrieved from both SAR and optical imagery by inverting canopy radiative transfer. For SAR imagery, the locally adjusted Water Cloud Model (WCM) is used and is inverted using a look-up table. The top volumetric soil moisture needed for the retrieval is evaluated using the Soil, Water, Atmosphere and Plant model (SWAP). To obtain LAI estimates in the solar domain, the Scattered Arbitrarily Inclined Leaves (SAIL) model is inverted using the neural network algorithm that was developed to produce the CYCLOPES global products (Baret et al. 2007). Overall, this intercomparison exercise provides a significant step toward a more holistic approach towards monitoring LAI dynamics at regional scale.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Bériaux, E., Defourny, P., & Duveiller Bogdan, G. (2009). Evaluating the potential of crop specific LAI time series retrieved from various SAR sensors to support agricultural monitoring at regional scale based on optical imagery. International GEO Workshop on Synthetic Aperture Radar (SAR) to Support Agricultural Monitoring, Kananaskis, Alberta, Canada. https://hdl.handle.net/2078.5/151419