Skip to main content

page search

Library Use of MODIS NDVI data to improve forest-area estimation

Use of MODIS NDVI data to improve forest-area estimation

Use of MODIS NDVI data to improve forest-area estimation

Resource information

Date of publication
December 2011
Resource Language
ISBN / Resource ID
AGRIS:US201400106842
Pages
6379-6393

The objective of this article is to develop and test a methodology capable of using medium spatial resolution satellite imagery to improve forest-area statistics derived from ground sampling. The methodology builds on the evidence that multitemporal Normalized Difference Vegetation Index (NDVI) images bring significant information on the spatial distribution of forest surfaces. Consequently, Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI images are potentially useful to improve forest-area assessment based on ground data. This expectation is verified in Tuscany (central Italy) using forest-area references extracted from the Coordination of Information on Environment (CORINE) land-cover map. The accuracy of forest-area statistics obtained at province level by different reference samplings is first assessed. Next, locally calibrated regression analyses are applied to multitemporal MODIS NDVI images in order to obtain per-pixel forest-area estimates. Two statistical methods (the direct expansion and the regression estimator) are finally used to combine these estimates with the ground data and produce corrected per-province statistics. The experimental results confirm that MODIS NDVI data contain relevant information on forest distribution, which can be efficiently extended over the land surface by locally calibrated regressions. The obtained estimates can be combined with the ground data for enhancing forest-area assessment at province level. To this aim, the regression estimator gives the best performance for all sampling densities of the reference data.

Share on RLBI navigator
NO

Authors and Publishers

Author(s), editor(s), contributor(s)

Maselli, Fabio

Publisher(s)
Data Provider
Geographical focus