Investigation of aggregation effects in vegetation condition monitoring at a national scale

Abstract

Monitoring vegetation condition is an important issue in the Mediterranean region, in terms of both securing food and preventing fires. Vegetation indices (VIs), mathematical transformations of reflectance bands, have played an important role in vegetation monitoring, as they depict the abundance and health of vegetation. Instead of storing raster VI maps, aggregated statistics can be derived and used in long-term monitoring. The aggregation schemes (zonations) used in Greece are the forest service units, the fire service units, and the administrative units. The purpose of this work was to explore the effect of the modifiable areal unit problem (MAUP) in vegetation condition monitoring at the above-mentioned aggregation schemes using 16 days Normalized Difference Vegetation Index (NDVI) composites acquired by the moderate resolution imaging spectroradiometer satellite sensor. The effects of aggregation in the context of MAUP were examined by analyzing variance, from which the among polygon variation (objects' heterogeneity) and the within polygon variation (pixels' homogeneity) were derived. Significant differences in objects' heterogeneity were observed when aggregating at the three aggregation schemes; therefore there is a MAUP effect in monitoring vegetation condition on a nationwide scale in Greece with NDVI. Monitoring using the fire service units has significantly higher pixels' homogeneity; therefore there is indication that it is the most appropriate for monitoring vegetation condition on a nationwide scale in Greece with NDVI. Results were consistent between the two major types of vegetation, natural and agricultural. According to the statistical validation, conclusions based on the examined years (2003 and 2004) are justified.

Keywords:

• modifiable areal unit problem
• zonation
• Vegetation Index
• vegetation monitoring
• MODIS

Acknowledgements

This study was funded by ‘Pythagoras’, a research grant awarded by the Managing Authority of the Operational Programme ‘Education and Initial Vocational Training’ of Greece, which is partially funded by the European Social Fund – European Commission. We are thankful to Ms. Stavroula Strati for her contribution to the image processing work. The Civil Protection Agency of Greece is gratefully acknowledged for its support. Authors are also thankful to the anonymous reviewers for their constructive comments on the original manuscript.