Skip to main content
ARS Home » Research » Publications at this Location » Publication #212096
Title: Fire severity modeling of sagebrush-steppe rangelands in Southeastern Idaho

Author
item WEBER, KEITH - IDAHO STATE UNIVERSITY
item Seefeldt, Steven
item NORTON, JILL - IDAHO STATE UNIVERSITY
item FINLEY, CHARLES - NEWMONT ENVIRONMENTAL

Submitted to: GIScience and Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/23/2008
Publication Date: 3/1/2008
Citation: Weber, K.W., Seefeldt, S.S., Norton, J.M., Finley, C. 2008. Fire severity modeling of sagebrush-steppe rangelands in Southeastern Idaho. GIScience and Remote Sensing. 45(1):68-82.

Interpretive Summary: Wildland fires are common in rangelands throughout the world and the sagebrush steppe ecosystem of the Intermountain West is no exception. Fires that burn with high severity have more negative impacts on vegetation than low severity fires. These high severity fires can result in long-term changes to rangeland vegetation. Some of these negative impacts can be corrected with management, but management success is dependent on quickly identifying high severity burn areas. Assessments are currently carried out in the field, which can be difficult in large and late season fires. To complement field methods, it may be possible to use satellite imagery to determine fire severity in rangelands. In this study, we applied post-fire field observations of SPOT satellite imagery using Classification Tree Analysis techniques at two 2005 burns in southeastern Idaho. The results of these analyses demonstrate that these types of analyses of SPOT satellite imagery are very good at classifying areas of low fire severity as well as those with high fire severity. Accuracy of the technique ranged from 66 to 100%. This technique provides a usable model determining fire severity for an entire fire area using readily available satellite imagery. Furthermore, this Classification Tree Analysis technique is fairly uncomplicated to apply and provides an error assessment upon which land managers can better justify their recommendations and decisions.

Technical Abstract: Wildland fires are common in rangelands throughout the world and the sagebrush steppe ecosystem of the Intermountain West is no exception. The potential for high severity fires to affect long-term changes in rangelands is considerable and for this reason, assessing fire severity shortly after the fire is critical. Such assessments are typically carried out following BAER or similar protocols. These field data can then be used by land managers to plan remediation and future land uses. To complement these procedures and explore fire severity modeling in rangelands, we applied post-fire field observations to SPOT satellite imagery using Classification Tree Analysis (CTA) techniques at two 2005 burns in southeastern Idaho. The results of these analyses demonstrate that CTA is a robust technique equally adept at classifying areas of low fire severity as well as those with high fire severity with reliable accuracy (user accuracy ranged between 66-100%). This technique provides a model for an entire fire area using readily available satellite imagery. Furthermore, the CTA technique is fairly uncomplicated to apply and provides an error assessment upon which land managers can better justify their recommendations and decisions.