Vegetation Monitoring of Palm Trees in an Oasis Environment (Boudenib, Morocco) Using Automatic Processing of Medium-Resolution Remotely Sensed Data

Oases are part of the natural wealth and heritage of Morocco and contribute to the social, economic, and touristic environment. Morocco has lost more than 2/3 of its oases during the past century due to water scarcity, succession of drought periods, climate change and over-exploitation of groundwater resources. Palm trees are strongly dependent on irrigation and availability of surface water as soon as the water table depth falls below the root zone of 9 m. Improving management and monitoring of oasis ecosystems is strongly encouraged by UNESCO Biosphere Reserve and RAMSAR guidelines. The Boudenib and Tafilalet oases are among the biggest palm groves located in the south-eastern part of Morocco. These oases belong to catchments of the rivers Guir and Ziz, respectively. This paper uses remotely sensed data from PROBA-V for monitoring vegetation in oases, and linking vegetation characteristics to water availability, water management and quality and quantity of date crops. The Normalized Differential Vegetation Index (NDVI) derived from optical images provides a good estimation of changes in vegetation cover over time. Images of various spatial resolutions (100 m, 300 m and 1 km) obtained with the frequently revisiting Belgian satellite PROBA-V and available since 2014, can be successfully used for deriving time series of vegetation dynamics. TREX—Tool for Raster data Exploration—is a Python-GDAL processing tool of PROBA-V NDVI images for analyzing vegetation dynamics, developed at the Vrije Universiteit Brussel and available online. TREX has various applications, but the main functionality is to provide an automatic processing of PROBA-V satellite images into time series of NDVI and LAI, used in vegetation monitoring of user-defined points of interest. This study presents the results of application of TREX in the arid ecosystems of the Boudenib oasis for the period 2014–2018. The resulting NDVI and LAI time series are also compared to time series of groundwater depth and date crops quantity and quality. Low LAI is observed when water depth is low, and the palm trees lose their greenery. Low LAI is also correlated to low quantity and quality of dates in October 2015 and October 2017. PROBA-V images can therefore be used for monitoring the health of palm trees in oasis environments. However, considering the fact that the PROBA-V satellite mission has ended, this approach could instead be applied to Sentinel-3 data using the same analysis. These results have important implications for water management in the area and can help decision-makers to make better decisions about prevention of water scarcity in the region.