Welcome to the InfoSci Platform
Reference Hub10
Using Spatial Statistics Tools on Remote-Sensing Data to Identify Fire Regime Linked with Savanna Vegetation Degradation

Using Spatial Statistics Tools on Remote-Sensing Data to Identify Fire Regime Linked with Savanna Vegetation Degradation

Anne Jacquin, Michel Goulard
Copyright: © 2013 |Volume: 4 |Issue: 1 |Pages: 15
ISSN: 1947-3192|EISSN: 1947-3206|EISBN13: 9781466631724|DOI: 10.4018/jaeis.2013010105
Cite Article Cite Article

MLA

Jacquin, Anne, and Michel Goulard. "Using Spatial Statistics Tools on Remote-Sensing Data to Identify Fire Regime Linked with Savanna Vegetation Degradation." IJAEIS vol.4, no.1 2013: pp.68-82. http://doi.org/10.4018/jaeis.2013010105

APA

Jacquin, A. & Goulard, M. (2013). Using Spatial Statistics Tools on Remote-Sensing Data to Identify Fire Regime Linked with Savanna Vegetation Degradation. International Journal of Agricultural and Environmental Information Systems (IJAEIS), 4(1), 68-82. http://doi.org/10.4018/jaeis.2013010105

Chicago

Jacquin, Anne, and Michel Goulard. "Using Spatial Statistics Tools on Remote-Sensing Data to Identify Fire Regime Linked with Savanna Vegetation Degradation," International Journal of Agricultural and Environmental Information Systems (IJAEIS) 4, no.1: 68-82. http://doi.org/10.4018/jaeis.2013010105

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

Fire is acknowledged to be a factor for explaining the disturbance of vegetation dynamics interacting with other environmental factors. In this study, the authors want to clarify the importance and the role of fire on the dynamics of savanna vegetation. The study area is the Marovoay watershed located on the north-west coast of Madagascar. In this site, burning herbaceous cover is the main practice in the extensive grazing system. They analyzed the relationship between two indicators, one related to vegetation activity changes and one about fire regime that results from a combination of fire frequency and seasonality. All indicators were measured between 2000 and 2007 using a time series of MODIS images. In this work, the authors implemented two approaches of spatial analysis. The first one analyzes the spatial structure of the residuals of a per-pixel non-spatial GLM model. In the second approach, a spatial GLM model is directly computed. In both approaches, the authors proposed two levels of stratification for the study area according to the spatial variations of the relationship established between vegetation activity changes and fire regime. The use of spatial statistical tools produces parsimonious models which they found to be consistent with expert knowledge. The authors demonstrated that a statistical analysis based on spatial GLM is able either to stratify an area when non ancillary data on land use exist or to validate an existing stratification.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.