The Correlation between Rainfall and Normalized Difference Vegetative Index in East Africa
Abstract
Vegetation dynamics, as measured by the Normalized Difference Vegetative Index (NDVI), a monthly rainfall are compared at seventy-eight stations in East Africa to determine spatial and temporal relationships between moisture and vegetation response. Rainfall and NDVI are found to be coincident in both space and time; with highest NDVI values persisting from year to year in areas of maximum rainfall. Large changes in monthly rainfall from one year to the next result in large NDVI changes. Due to the spatial variability of vegetation types in East Africa and the varying NDVI response of vegetation to moisture, correlations of monthly-integrated NDVI and monthly performed by plant zone. To determine the time lag between rainfall and resulting NDVI change, NDVI is regressed against averages of rainfall of previous months. The combination giving the highest coefficient of correlation is used to quantify the time large between rain and NDVI response. Monthly-integrated NDVI and monthly rainfall of the stations within a single plant zone correlate well. When the three-year averages of yearly-integrated NDVI and yearly rainfall are compared, a strong relationship is obtained (r-.76). This relationship could provide a means of estimating annual rainfall over data sparse regions. An improved vegetative mapping scheme is devised and applied to the three-year average of the data. The mapping scheme results in a mapping that is far superior to previous mappings using NDVI data. Keywords: Theses.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1988
- Accession Number
- ADA196292
Entities
People
- Michael L. Davenport
Organizations
- Air Force Institute of Technology