Detection of seasonal land use pattern and irrigated crops in drylands using multi-temporal sentinel images

The objective of this research is to detect seasonal agricultural land use and cropping pattern of irrigated crops during the 2017/2018 growing season in the Mujib basin in Jordan. The Mujib basin is considered as one of the major sources for water production in Jordan, while the agricultural production is the main consumer sector for water in the basin. High multi-temporal multi-spectral resolutions Sentinel-2b satellite images were used to identify the land use types using the supervised classification that applied on spring and summer growing season. The results show that the temporal NDVI curve is very useful in distinguishing between different crops type with high details, and is used in assisting the agricultural land use pattern. Multi-temporal NDVI images were used to identify the cropping pattern within the agricultural land using the unsupervised classification approach. The results show that the rainfed agriculture is confined in the high rainfall zones with one period of growing season, while the irrigated agricultural crops are spreading in the basin over the whole year with different periods of growing seasons. Irrigated vegetables and tree crops are the dominant agricultural crops in the basin with more than 70% of cultivated land. Most of the irrigated vegetables and trees are grown in the summer seasons under high temperatures conditions which consumes the highest percentage of irrigated water. These results show that the patterns of agricultural land use under Dryland regions are very essential issue in determining water consumption and management alternatives.