Super typhoon Haiyan makes landfall
Super typhoon Haiyan
Subscribe here to receive special images

VEGETATION - 2000

Lake Maggiore - Italy, 3-6 April 2000


VEGETATION potentialities in food early warning systems in the Sahelian region

TYCHON Bernard , OZER Pierre and TOURE Souleymane
Fondation Universitaire Luxembourgeoise
185, Avenue de Longwy, B-6700 Arlon - Belgium
Email : tychon@ful.ac.be

Paper (pdf file, 316 k)

Developing countries in the Sahelian region dedicate most of their activities to Agriculture. A year with rainfall deficits or unevenly distributed rainfall during the rainy season often leads to food insecurity in this part of the world. This food status has to be foreseen early enough to allow decision-makers or financial backers to react fast enough to prevent or reduce the effects of potential famines. Among the available tools to foresee this type of disaster, the combined use of the low resolution NOAA-AVHRR sensor with the IR band of METEOSAT has served up to now in the global vegetation monitoring in Africa. This sensor association is presently used in routine by FAO inside its ARTEMIS Programme. However, the AVHRR sensor has shown some limitations that did not always allow reaching the initial expectation of such an earth observation sensor. Some of these limits have been reduced even removed with VEGETATION sensor placed on SPOT 4.

The general objective of this study is to verify that VEGETATION instrument actually provides information of better quality than this derived from NOAA satellite specifically for the Sahelian region and within the scope of early warning system on food and agriculture of FAO.

Two North-South transects were chosen inside the Sahelian region as study-area: the first one from the Malian Sahel to southern Burkina Faso; the second one inside Niger only concerns the Sahelian band.

Four criteria were selected to compare remote observations from the two satellites:

The first criterion is based on the correlation between local yields of millet calculated with an agrometeorological model (DHC-CP) and the NDVI from NOAA and VGT.

The second criterion checks the potentiality of spatial extrapolation of agrometeorological parameters. Here, the chosen criterion allows determining the start of the vegetation season with remote sensing based on a set of image series. Results from the two remote sources are compared with a map calculated with an agrometerological approach that fixes sowing date according to a given quantity of rainfall per 10-day period.

A third criterion checked the saturation level of both sensors in regions with high vegetation density.

Finally, pixel location was analysed inside a window crossed by the Niger River.

The VEGETATION instrument demonstrated its higher capacities for vegetation monitoring in the Sahelian region within the scope of agricultural campaign monitoring and food early warning systems in comparison with the NOAA-AVHRR sensor presently used for this topic.

These potentialities should now be used to replace NOAA images in the different monitoring systems. Moreover, we think that VGT should be used for other applications than those presently run by NOAA. Especially, it should be necessary to look at the potentialities of the sensor in the monitoring of small areas (< 100 kmĀ²). It should also be worth improving the integration of remote sensing information inside agrometeorological models in order to valorise this new type of information in a quantitative way.