Sitemap Contact Kalender Nieuw Home

Projectendatabank FEDRA





Dynamic predictive mapping using multi-sensor data fusion – demonstration for malaria vector habitat (DYNMAP)

Onderzoeksproject SR/00/107 (Onderzoeksactie SR)

Personen :

  • Prof. dr.  DEFOURNY Pierre - Université Catholique de Louvain (UCL)
    Coördinator van het project
    Betoelaagde Belgische partner
    Duur: 1/12/2006-30/11/2008
  • Prof. dr.  COOSEMANS Marc - Prins Leopold Instituut voor Tropische Geneeskunde (ITG)
    Betoelaagde Belgische partner
    Duur: 1/12/2006-30/11/2008
  • Dr.  BOGAERT Patrick - Université Catholique de Louvain (UCL)
    Betoelaagde Belgische partner
    Duur: 1/12/2006-30/11/2008

Beschrijving :

Context and objectives

This project aims at developing statistically-sound methods to update land surface descriptors in a near-automatic way. While high resolution EO systems provide now very regular update of the earth surface state, very few methods allow integrating the data flow into existing land products. Up-to-date information is however crucial for the monitoring of environment-related processes. Innovative methods are thus needed to provide more frequent updates and increase the opportunity of dynamic predictive mapping. Multi-sensor data fusion and downscaling techniques combined to statistical modelling could offer an alternative.

This research has four objectives:

1. Investigate how Bayesian data fusion can derive high resolution reflectance values based on medium resolution observations using a multivariate approach including covariate information, and thus develop a new Bayesian data fusion approach merging multiple data source of different spatial resolution referring to the same timeframe.
2. Adapt this new approach to temporally non overlapping dataset.
3. Use this method to update in a near-automatic way land surface descriptors of interest for malaria vector control in Southeast Asia: land cover descriptor and relative humidity proxies fusion and dynamic mapping
4. Analyse the relation between land descriptors and malaria vectors to assess performance of the method in an application context. This will include modelling vector occurrence using land cover and relative humidity descriptors and finally the delineation of restricted zone for dry season vector habitat


• Develop a new Bayesian data fusion approach (BDF) merging multiple data sources of different spatial resolution referring to the same timeframe and adapt it to temporally non overlapping dataset.
• Implementation on land descriptors for one study site and validation on a second. The steps include (1) Fusion of panchromatic and multispectral images of high resolution SPOT5 HRV images (2) Fusion of low/medium resolution (MODIS TERRA and ENVISAT-Meris) and high resolution imagery (3) Use medium resolution image at time T to study the relation between the high resolution image and the medium resolution images (4) Use medium resolution images available throughout a year to derive updated information on the framework of the high resolution image (5) Use of high resolution image for T+1and T+2 to validate the result.
• Implementation on water content descriptors such as leaf water content information and other humidity proxies in forested areas (habitat of the main malaria vector of the region: An. dirus)
• Interpretation for An. dirus primary sites and validation

Results expected

• A new data fusion method applicable not only to concomitant images of various resolutions but also to update detailed images using low to medium resolution imagery
• Improved knowledge of relation between land descriptors and malaria vector in South East Asia
• Verification of the hypothesis of receding habitat in the dry season for Anopheles dirus s.l. and association with land cover and relative humidity
• Validation of leaf water content as a valid indicator for approximation of relative humidity related to mosquito habitat
• Pave the way towards a new family of products usable for various application
• To help in improving the use of remote sensing product in the field of epidemiology

• Peer-reviewed scientific articles describing the achievements, limitation and applicability of the new developed methodology
• The activities and results of the project will be compiled in reports. A particular attention will be made to describe recommendations and basis for the development of a decision support tool
• Map predicting the habitat of the vector in the dry season if the hypothesis is validated

Documentatie :

Over deze website


© 2018 POD Wetenschapsbeleid