Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.539846
Title: Influence of land use, climate and topography on the fire regime in the Eastern Savannas of Colombia
Author: Ruiz, Milton Hernan Romero
Awarding Body: University of Leicester
Current Institution: University of Leicester
Date of Award: 2011
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
Changes in natural vegetation as a result of land use and fire are an important factor that contributes to loss of biodiversity and climatic change. More than 30% of the global land surface shows frequent burning particularly in the tropics. Even though natural savannas in Colombia are experiencing high impact due to land use, there has not been a systematic quantification of the change rate of land use/cover or fire occurrence. The Eastern Colombian Savannas represent around 6% of the savannas in South America. This study identifies the land use/ cover change (LUCC) patterns between three periods (1987-1988, 2000-2001 and 2006-2007) using Landsat and CBERS satellite images. The area burned between 2000-2009 using a novel regional algorithm tailored for MODIS data is quantified and validated. Results show that for the year 2000, 22% of the savanna ecosystems had been transformed, with flooded and high savannas being most affected. The annual rate of change from natural to non-natural between 1987- 2007 was -0.85%. The fire assessment showed that on average 2.75 +/- 0.5 million ha of the savannas burn each year, being highly variable as 3.4 million ha burned in 2002-2003 which was 1.9 million ha less than in 2005-2006. However, it was shown that during 2000-2009, 39% of the savannas never burned. Fires predominate in the high plain savannas, with lowest occurrence along the Andean foothills, in forested areas and croplands. Based on predictive models implementing statistical methods, it was shown that in the occurrence of fires, the climate variables are important but only predict a 44%. The best predictive model with an accuracy of 78% includes variables such as climate, land use/ cover, land form and topography. This study presents the first complete map at regional scale for almost two decades of LUCC and a decade of fire activity in Colombian savannas.
Supervisor: Tansey, Kevin ; Berrio, Juan Carlos Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.539846  DOI: Not available
Share: