Gabriele Caccamo

Monitoring live fuel moisture content of heathland, shrubland and sclerophyll forest in south-eastern Australia using MODIS data

Live fuel moisture content is an important variable for assessing fire risk. Satellite observations provide the potential for monitoring fuel moisture across large areas. The objective of this study was to use data from the Moderate Resolution Imaging Spectroradiometer to monitor live fuel moisture content of three fire-prone vegetation types (shrubland, heathland and sclerophyll forest) in south-eastern Australia. The performances of four spectral indices (Normalised Difference Vegetation Index, Visible Atmospherically Resistant Index, Normalised Difference Infrared Index centred on 1650 nm and Normalised Difference Water Index) were compared. Models based on Visible Atmospherically Resistant Index and Normalised Difference Infrared Index centred on 1650 nm provided the best results (R2 values of 0.537 and 0.586). An empirical model based on these two indices was developed and its performance compared with a meteorological index traditionally used in this context, the Keetch–Byram Drought Index. The empirical model (R2 = 0.692) outperformed the meteorological index (R2 = 0.151), showing an enhanced capability to predict live fuel moisture content of the fire-prone vegetation types considered.

Large-scale, dynamic transformations in fuel moisture drive wildfire activity across southeastern Australia

The occurrence of large, high-intensity wildfires requires plant biomass, or fuel, that is sufficiently dry to burn. This poses the question, what is “sufficiently dry”? Until recently, the ability to address this question has been constrained by the spatiotemporal scale of available methods to monitor the moisture contents of both dead and live fuels. Here we take advantage of recent developments in macroscale monitoring of fuel moisture through a combination of remote sensing and climatic modeling. We show there are clear thresholds of fuel moisture content associated with the occurrence of wildfires in forests and woodlands. Furthermore, we show that transformations in fuel moisture conditions across these thresholds can occur rapidly, within a month. Both the approach presented here, and our findings, can be immediately applied and may greatly improve fire risk assessments in forests and woodlands globally.

Using MODIS data to analyse post-fire vegetation recovery in Australian eucalypt forests

Remote sensing observations provide useful spatially explicit and temporally dense information for monitoring post-fire vegetation recovery patterns over large areas. Although large fires are common in Australian eucalypt forests, research on remote sensing of post-fire vegetation recovery in this ecosystem has been limited. In this study, time series (2000–2012) of Normalised Difference Vegetation Index, Enhanced Vegetation Index and Normalised Differenced Infrared Index derived from Moderate Resolution Imaging Spectroradiometer (MODIS) were used to analyse post-fire vegetation recovery in eucalypt forests in Australia. The analysis focused on 11 sites which burned during 2001/02 and 2002/03 fire seasons. Results indicated that spectral recovery in Australian eucalypt forests is particularly rapid after fire as spectral indices values returned to pre-fire levels three to six years after fire. Spectral recovery was particularly rapid during the first year following fire and the influence of severity was limited to the first two years after fire.