Space and RFV in Senegal

SPOT-5 imagery allowed ponds detection in the Ferlo region of the Sahel using high-resolution ( 10 m pixels) images. Four spectral bands/channels were used, i.e. the green channel (0.50 to 0.59 μm), the red channel (0.61 to 0.68 μm), the near-infrared channel (NIR, 0.78 to 0.89 μm) and the mediuminfrared channel (MIR, 1.58 to 1.75 μm). Radiometric and geographic corrections, needed due to small changes in satellite altitude, inclination, position relative to the sun, curvature of earth, state of the troposphere, sensor calibration, were implemented by SPOT-Image. Digital counts (DCs) from radiometric values within each pixel, and “DC thresholds” from photo-interpretation were used for class ranking (for details, see Lacaux et al.,2007).

The classification for each pixel was obtained by using a decision-tree classifier distinguishing between water and other soil elements. As such, new indices were developed, i.e. the normalized difference pond index (NDPI) and the normalized difference turbidity index (NDTI). Lacaux et al.(2007) provide details on how NDPI and NDTIwere computed.

Data could be acquired at least every 26 days(orbital cycles of polar- and helio-synchronous satellites), with a revisit time of 3 days, assuming minimum cloud coverage, allowing detailed spatio-temporal evolution and dynamics of the ponds. The programmed five SPOT-5 images (60 x 60 km each) were centred near the Barkedji village in the Ferlo , covering the 2003 rainy season in its entirety.

This area is particularly suitable for the study since detailed entomological data, collected and analyzed previously (see above), contribute much to the understanding of the dynamics of the ponds considered as “vector producers”. In addition, specific information on pond characteristics is available in Lacaux et al. (2007) and Tourre et al. (2008). Among other parameters, the Tropical Rainfall Measuring Mission (TRMM) identifies rainfall events and estimates their amounts of water. TRMM, which has been in operation since January 1998, is a joint space mission between NASA and the Japan Aerospace Exploration Agency (JAXA) designed to specifically monitor tropical rainfalls.

The resulting data were downloaded from the NASA website and averaged over the Barkedji area. As a working hypothesis, when few consecutive rainy days occurred, they were assimilated into one rainfall event (with the corresponding total rainfall amount). As discussed above, the presence and density of Ae. vexans mosquitoes are known to be governed by the frequency and intensity of rainfall as these events modify both the extent and the distribution of the ponds. The combined approach of ground checking and satellite monitoring takes into account not only the total amount of rainfall, but also the true accumulation of water within ponds and their extensions, all key parameters for mosquito production and ZPOM mapping (Vignolles et al. (2009)).

Data from the QuickBird satellite, a high-resolution commercial earth-observation satellite launched in 2001, was used for geo-referencing only. Its panchromatic imagery is collected with a 0.6 m resolution and multi-spectral imagery at 2.4 m resolution. Details of structures such as areas fenced-in for animal keeping are thus easily visible with precise coordinates. The latter, embedded into the August 26, 2003 SPOT-5 image, allowed a precise mapping of the study area at that date. As a result, ZPOM mapping with pond dynamics includes all fenced-in areas where vulnerable animals were kept at night in the Barkedji area.