Analysis of vegetation and land cover dynamics in north-western Morocco during the last decade using MODIS NDVI time series data

摘要: Abstract. Vegetation phenology as well the current variability and dynamics of vegetation land cover, including its climatic human drivers, are examined in a region north-western Morocco that is nearly 22 700 km2 big. A gapless time series Normalized Differenced Index (NDVI) composite raster data from 29 September 2000 to 2009 utilised. The have spatial resolution 250 m were acquired by Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. presented approach allows compose analyse yearly cover maps widely unknown with scarce validated ground truth deriving phenological parameters. Results show high temporal 16 d sufficient for (a) determining local better than global classifications Plant Functional Types (PFT) Global Land Cover (GLC2000) (b) drawing conclusions on drivers. Areas stably classified types (i.e. areas did not change their type) climatically driven inter- intra-annual indicated influence droughts. determine human-driven caused agriculture results more ten times larger area compared areas. Change detection based shows gain high-productive (cropland) about 259.3 km2. Statistically significant inter-annual trends during last decade could however be discovered. sequence correlations was respectively carried out extract most important periods rainfall responsible production green biomass extent types. mean daily precipitation 1 October 15 December has correlation (max. r2=0.85) an scale NDVI percentiles (50 %) Correlation January percentage each type medium up r2=0.64). In all, offset 1.5 months detected between rates values. High-productive proved mainly rain-fed. We conclude identification, understanding knowledge phenology, prediction methods change, can improved using multi-year MODIS data. This study enhances comprehension surface Morocco. It especially offers quick access when estimating agricultural lands.