CLASSIFICATION OF MULT-SSPECTRAL DATA TO DISTIGUISH PLANT HEALTHY STATUS(TEA)

This research aimed at comparing tea healthy status of satemwa tea field. Earth observation data downloaded on united states geological survey has been used which include Landsat 7 images from which Land surface temperature, Normalized difference vegetation index, soil adjusted vegetation index, spectral signatures and classified maps were generated, Study area polygon obtained on google earth which helped to delineate the study area, soil and precipitation data obtained on cru. Data website and earth map respectively. Training areas from the Landsat 7 images were generated using supervised approach to come up with spectral signature plots which has indicate decrease in reflectance of tea plants across the study period. NDVI, SAVI and Land surface temperature results has also shown decrease across the study periods. Results of annual precipitation shows increase across the study period. The results of the research would prompt farm managers to go into tea fields to see what has contributed to decrease in healthy status in the case that the biophysical factors under assessment shows little variation. For the purpose of this study only Land surface temperature, soil and precipitation were assessed but the methodologies can be simulated to assess other factors such as humidity and moisture.