Vegetation types and patterns in Senegal based on multivariate analysis of field and NOAA‐AVHRR satellite data

Abstract

Abstract. Woody vegetation was studied in 44 sites in Senegal. Species composition and density were analysed with TWINSPAN and Canonical Correspondence Analysis; 16 vegetation types were defined. A supervised, multispectral and multitemporal classification of day and night, dry season NOAA‐AVHRR imagery was used to identify their distribution with a classification accuracy ranging from 60–100%.