Detection and mapping of long-term land degradation and desertification in Arab region using modeler

الملخص EN

This paper presents the definition and formulation of a remotely sensed based monitoring system for land degradation and desertification.

The main objective of this system is to identify processes of land degradation and desertification sensitive areas at regional scale.

The proposed methodology relies on the use of time series NDVI images generated from remote sensing satellites for the regional assessment and monitoring of vegetation dynamics.

Parametric model based on ordinary least-squares regression was developed and implemented through the use of MODESERT, an open-source software developed by the National Council for Scientific Research in Lebanon (CNRS).

The study area covers the entire Arab World.

Long term trends were derived using the GIMMS NDVI data set over the time period between 1982 and 2006.

It is the longest series of NDVI product covering our regions.

Three main classes, subdivided into seven groups, were established representing: moderately to highly sensitive areas, areas with no-significant change, moderately to highly developed areas.

Mean absolute percentage error (MAPE) was used to estimate the accuracy of the trend model; it generated a mean error at 5-10%.

Results indicated that more than 40% of the total MENA region were sensitive to land degradation and desertification.

In contrast, only less than 5% of the region had witnessed positive changes in vegetation cover.

The major drivers of land degradation were mainly linked to climate variability and recurring drought and to less extent their sub-drivers such as forest fires and the poor socioeconomic development.

Development in vegetation cover was found to occur at larger scales in Southern Somalia, Sudan and Southern Iraq marshlands.

This method could be applied to national scale assessment.

Using others satellites images (e.g.

SPOT Vegetation, MOD13Q1, etc…) has also potential.

This monitoring system is operational since 2007 and was validated through the publication of the biennial Arab Desertification Bulletin.