The Vegetation Structure Perpendicular Index (VSPI): a forest condition index for wildfire predictions

Andrea Massetti, Christoph Rüdiger, Marta Yebra, James Hilton

Research output: Contribution to journalArticleResearchpeer-review

28 Citations (Scopus)


Wildfires are a major natural hazard, causing substantial damage to infrastructure as well as being a risk to lives and homes. An understanding of their progression and behaviour is necessary to reduce risks and to develop operational management strategies in the event of an active fire. Many empirical fire-spread models have been developed to predict the spread and overall behaviour of a wildfire, based on a range of parameters such as weather and fuel conditions. However, these parameters may not be available with sufficient accuracy or spatiotemporal resolution to provide reliable fire spread predictions. Fuel condition data include variables such as vegetation quantity, structure and moisture content and, in the event of previous wildfires, the burn severity and stage of ecosystem recovery. In this study, an index called the Vegetation Structure Perpendicular Index (VSPI) is introduced. The VSPI utilises the short-wave infrared reflectance in bands centred at 1.6 and 2.2 μm, essentially representing the amount and structure of the vegetation's woody biomass (as opposed to the photosynthetic activity and moisture content). The VSPI is quantified as the divergence from a linear regression between the two bands in a time series and represents vegetation disturbance and recovery more reliably than indices such as the Normalised Burn Ratio (NBR) and Normalised Difference Vegetation Index (NDVI). The VSPI index generally shows minor inter-annual variability and stronger post-wildfire detection of disturbance over a longer period than NBR and NDVI. The index is developed and applied to major wildfire events within eucalypt forests throughout southern Australia to estimate both burn severity and time to recovery. The VSPI can provide an improved information layer for fire risk evaluation and operational predictions of wildfire behaviour.

Original languageEnglish
Pages (from-to)167-181
Number of pages15
JournalRemote Sensing of Environment
Publication statusPublished - 1 Apr 2019


  • Forest condition
  • Landsat
  • Vegetation recovery
  • Wildfire spread

Cite this