Identifying Relationships Between The Vegetation Cover And Climatic Characteristics Using NOAA AVHRR

Abstract

Identifying Relationships Between The Vegetation Cover And Climatic Characteristics Using NOAA AVHRR

N.D.K. Dayawansa, Ranjith Premalal De Silva and S.M.M.S.K. Sooriyakula
Department of Agricultural Engineering
Faculty of Agriculture, University of Peradeniya, Sri Lanka
Tel: 0094-81-288923, Fax: 0094-81-280125
Email: rpdesilva@pdn.ac.lk

Abstract
Status of the vegetation cover is a good indicator of the prevailing climatic conditions of a particular area. Satellite remote sensing facilitates the identification of vegetation cover and the changes over time very effectively compared to the conventional field surveying techniques. This study focussed on a methodology to identify relationships between the vegetation cover and the climatic variables (rainfall and land surface temperature) using low resolution public domain satellite data.

Monthly maximum cloud-free NOAA AVHRR LAC lB data were downloaded from the NOAA online data archives for this study. For the months where it was difficult to obtain single cloud free images, mosaic images were produced with the cloud free parts obtained from the adjacent dates. Normalized Difference Vegetation Index (NDVI) values were derived for each month using the image mosaics. Monthly Land Surface Temperature (LST) images were derived using the equation adopted by Uliveri et al., (1994). Monthly rainfall data obtained from 385 rain gauging stations over the country was used to generate the monthly rainfall surfaces. In identifying the relationships between vegetation cover and climatic variables, it was not possible to perform the analysis for the entire country due to the presence of clouds in some parts of the images. Therefore, random samples were selected from the major climatic zones namely, Wet Zone, Intermediate Zone, Dry Zone, and Semi Arid Zone for the analysis. The correlation between the NDVI and the one month lag rainfall was analysed using regression analysis for the selected samples. The relationship between the LST and the occurrence of inter-monsoonal rains was also investigated.

The results demonstrated a varying degree of correlation and probability values of regression. Specifically, the correlation between vegetation cover and rainfall in Dry, Semi Arid and Intermediate Zones were the strongest. In Wet Zone, there was no strong relationship. Significant LST variations were identified during monsoonal and convective rainfall periods in different climatic zones. High LST values were observed during the convective periods. This study was limited to a single year and it is recommended to carry out the analysis for a longer time series in order to establish more stable relationships between vegetation cover and climatic variables. These relationships can be used in filling the gaps in rainfall time series at a location with the vegetation cover and LST information is available ffrom satellite data.