Author： Tanase Mihai de la Riva Juan Pérez-Cabello Fernando
Publisher： NRC Research Press
ISSN： 1208-6037
Source： Canadian Journal of Forest Research, Vol.41, Iss.4, 2011-04, pp. : 863-872
Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.
Abstract
During the last decades, the average number of fires per year increased significantly. A twofold increase was observed in the Mediterranean Basin, whereas in the western United States, the increase was fourfold. Regional models for burn severity estimation are necessary to avoid time consuming and costly fieldwork at each individual site. Furthermore, the estimation errors should be assessed by burn severity classes to avoid overestimating models accuracy. To develop such models, this study assessed the relationship between the composite burn index (CBI) and several spectral indices across five burned sites in northeastern Spain. The nonlinear models coupled with spectral indices containing information from the short wavelength infrared provided the best statistical fit of the data at most individual sites and for the pooled data set. The estimation errors for highly burned sites were well below 10%, but for burned sites of low and moderate severity, the errors increased significantly. A strong linear relation was found between burn severity at the plot level and understory and overstory composites. This study demonstrates (
Related content
Access to resources
Recommend
