---

The main objective of the study was to establish the spatial structure of Yala swamp, determine plant community distribution and wetland landscape characteristics in relation to other swamp characteristics such as main channels and water bodies. The study employed remote sensing techniques of computer satellite image processing to derive and map spatial land cover types of Yala swamp using LANDSAT TM. The satellite imagery used as primary data source was from LANDSAT TM recorded on the 22nd March 2000. Classification scheme based on the spatial distribution of plant communities and landscape features was developed. The satellite image was classified on the basis of a modified Anderson et al (1979) land cover classification scheme. Digital image processing was done in IDRISI (ver.1.0) environment, Classification accuracy assessment was undertaken using confusion matrix to determine the degree of fit between the classified and the reference data or ground data.

---

Nyunja Otieno, Maurice
Maurice Nyunja Otieno graduated from Moi University.

---