Multitemporala vegetationsstudier i nordöstra Kenya med AVHRR NDVI

Sammanfattning: In this paper coarse spatial resolution, high temporal frequency satellite data from the NOAA/AVHRR system are presented and analysed to demonstrate their utility for monitoring vegetation in a semiarid environment and for describing its seasonal dynamics. The study area consists of four districts in the North East of Kenya, namely Marsabit, Samburu, Mandera and Wajir. Normalized Difference Vegetation Index (NDVI) data have been derived and analysed for the major vegetation types from two years, 1984 (an unusually dry period) and 1990 (normal precipitation). The best vegetation class separability was found in the 1984 data, where the woody vegetation retained its spectral properties during the dry period. Within a single year the greatest separability is found during the rainy seasons. Rainfall data was used to derive a relationship between NDVI and precipitation in the semiarid areas. The study shows that events with rainfall mainly control the phenology in the area.