Remote sensing of fire severity in the Blue Mountains: influence of vegetation type and inferring fire intensity
Kate A. Hammill A B and Ross A. Bradstock AA Biodiversity Conservation Science Section, Department of Environment and Conservation (NSW), PO Box 1967, Hurstville, NSW 2220, Australia.
B Corresponding author. Email: kate.hammill@environment.nsw.gov.au
International Journal of Wildland Fire 15(2) 213-226 https://doi.org/10.1071/WF05051
Submitted: 27 April 2005 Accepted: 2 March 2006 Published: 31 May 2006
Abstract
Fire intensity affects ecological and geophysical processes in fire-prone landscapes. We examined the potential for satellite imagery (Satellite Pour l’Observation de la Terre [SPOT2] and Landsat7) to detect and map fire severity patterns in a rugged landscape with variable vegetation near Sydney, Australia. A post-fire, vegetation-based indicator of fire intensity (burnt shrub branch tip diameters, representing the size of fuel consumed) was also used to explore whether fire severity patterns can be used to retrospectively infer patterns of fire intensity. Six severity classes (ranging from unburnt to complete crown consumption) were defined using aerial photograph interpretation and a field assessment across five vegetation types of varying height and complexity (sedge-swamp, heath, woodland, open forest, and tall forest). Using established Normalised Difference Vegetation Index (NDVI) differencing methodology, SPOT2 and Landsat7 imagery yielded similar broad-scale severity patterns across the study area. This was despite differences in image resolution (10 m and 30 m, respectively) and capture dates (2 months and 9 months apart, respectively). However, differences in the total areas mapped for some severity classes were found. In particular, there was reduced differentiation between unburnt and low-severity areas and between crown-scorched and crown-consumed areas when using the Landsat7 data. These differences were caused by fine-scale classification anomalies and were most likely associated with seasonal differences in vegetation condition (associated with time of image capture), post-fire movement of ash, resprouting of vegetation, and low sun elevation. Relationships between field severity class and NDVIdifference values revealed that vegetation type does influence the detection of fire severity using these types of satellite data: regression slopes were greater for woodland, forest, and tall forest data than for sedge-swamp and heath data. The effect of vegetation type on areas mapped in each fire severity class was examined but found to be minimal in the present study due to the uneven distribution of vegetation types in the study area (woodland and open forest cover 86% of the landscape). Field observations of burnt shrub branch tips, which were used as a surrogate for fire intensity, revealed that relationships between fire severity and fire intensity are confounded by vegetation type (mainly height). A method for inferring fire intensity from remotely sensed patterns of fire severity was proposed in which patterns of fire severity and vegetation type are combined.
Alexander ME (1982) Calculating and interpreting forest fire intensities. Canadian Journal of Botany 60, 349–357.
Catling PC, Coops NC , Burt RJ (2001) The distribution and abundance of ground-dwelling mammals in relation to time since wildfire and vegetation structure in south-eastern Australia. Wildlife Research 28, 555–564.
| Crossref | GoogleScholarGoogle Scholar |
Clarke PJ (2002) Habitat islands in fire-prone vegetation: do landscape features influence community composition. Journal of Biogeography 29, 677–684.
| Crossref | GoogleScholarGoogle Scholar |
Franklin SE, Hall RJ, Smith L , Gerylo GR (2003) Discrimination of conifer height, age and crown closure classes using Landsat-5 TM imagery in the Canadian Northwest Territories. International Journal of Remote Sensing 24, 1823–1834.
| Crossref | GoogleScholarGoogle Scholar |
Gill AM, Allan G , Yates C (2003) Fire-created patchiness in Australian savannas. International Journal of Wildland Fire 12, 323–331.
| Crossref | GoogleScholarGoogle Scholar |
Milne AK (1986) The use of remote sensing in mapping and monitoring vegetational change associated with bushfires events in eastern Australia. Geocarto International 1, 25–32.
Price O, Russell-Smith J , Edwards A (2003) Fine-scale patchiness of different fire intensities in sandstone heath in northern Australia. International Journal of Wildland Fire 12, 227–236.
| Crossref | GoogleScholarGoogle Scholar |
Turner MG, Hargrove WW, Gardner RH , Romme WH (1994) Effects of fire on landscape heterogeneity in Yellowstone National Park, Wyoming. Journal of Vegetation Science 5, 731–742.
| Crossref |
Whight S , Bradstock RA (1999) Indices of fire characteristics in sandstone heath near Sydney, Australia. International Journal of Wildland Fire 9, 145–153.
| Crossref | GoogleScholarGoogle Scholar |