Evaluating the near and mid infrared bi-spectral space for assessing fire severity and comparison with the differenced normalized burn ratio

Max Van Gerrevink,Sander Veraverbeke

doi:10.5194/egusphere-egu21-2238

Abstract

&lt;p&gt;Fire severity, defined as the degree of environmental change caused by a fire, is a critical fire regime attribute of interest to fire emissions modelling and post-fire rehabilitation planning. Remotely sensed fire severity is traditionally assessed by the differenced normalized burned ratio (dNBR). This spectral index captures fire-induced reflectance changes in the near infrared (NIR) and short-wave infrared (SWIR) spectral regions. This study evaluates a spectral index based on a band combination including the NIR and mid infrared (MIR) spectral regions, the differenced normalized difference vegetation index (dNDVI&lt;sub&gt;MID&lt;/sub&gt;), to assess fire severity. This evaluation capitalized upon the unique opportunity stemming from the pre- and post-fire airborne acquisitions over the Rim (2013) and King (2014) fires in California with the MODIS/ASTER (MASTER) instrument. The field data consists of 85 Geometrically structured Composite Burn Index (GeoCBI) plots. In addition, six different index combinations, respectively three with a NIR-SWIR combination and three with a NIR-MIR combination, were evaluated based on the optimality of fire-induced spectral displacements. The optimality statistic ranges between zero and one, with values of one representing pixel displacements that are unaffected by noise. Results show that the dNBR demonstrated a stronger relationship with GeoCBI field data when field measurements over the two fire scars were combined than the dNDVI&lt;sub&gt;MID&lt;/sub&gt; approaches. The results yielded an R&lt;sup&gt;2&lt;/sup&gt; of 0.68 based on a saturated growth model for the best performing dNBR index, whereas the performance of the dNDVI&lt;sub&gt;MID &lt;/sub&gt;indices was clearly lower with an R&lt;sup&gt;2&lt;/sup&gt; = 0.61 for the best performing dNDVI&lt;sub&gt;MID &lt;/sub&gt;index. The dNBR also outperformed the dNDVI&lt;sub&gt;MID&lt;/sub&gt; in terms of spectral optimality across both fires. The best performing dNBR index yielded the optimality statistics of 0.56 over the Rim and 0.60 over the King fire. The best performing dNDVI&lt;sub&gt;MID, &lt;/sub&gt;index recorded optimality values of 0.49 over the Rim and 0.46 over the King fire. We also found that the dNBR approach led to considerable differences in the form of the relationship with the GeoCBI between the two fires, whereas the dNDVI&lt;sub&gt;MID&lt;/sub&gt; approach yielded comparable relationships with the GeoCBI over the two fires. This suggests that the dNDVI&lt;sub&gt;MID&lt;/sub&gt;&amp;#160;approach, despite its slightly lower performance than the dNBR, may be a more robust method for estimating and comparing fire severity over large regions. This premise needs additional verification when more air- or spaceborne imagery with NIR and MIR bands will become available with a spatial resolution that allows ground truthing of fire severity.&amp;#160;&lt;/p&gt;

Highlights

For the differenced normalized burn ratio (dNBR) indices, the relationship with field data was stronger and different across the Rim and King fire (Table 4). This difference in the form of the regression line is less pronounced in the relationships between the two fires for the dNDVIMID indices (Figure 5d–f)
The results of our comparison between field and remotely sensed proxies of fire severity and the spectral optimality analysis confirmed the dNBR as a strong fire severity predictor
The field data analysis, identified that the dNBR approach suffered from significant variations in the form of regression lines across the two fires

Summary

Introduction

Wildfires in the Western United States have intensified during the recent decades with respect to their size, severity, and frequency [1–5]. Analysing the effects of fires on ecosystems by assessing how fires impact vegetation recovery and succession is critical to forest management [6,7]. We adopt the fire continuum framework of Jain [14] to distinguish between fire and burn severity In this framework, fire severity quantifies the short-term fire effects on the immediate post-fire landscape, and as such, it mainly quantifies vegetation consumption and soil alteration [15,16]. Burn severity assesses and quantifies both the short-term and long-term effects by including longer-term ecosystem response processes such as delayed tree mortality and vegetation recovery [15,17]. We use the term fire severity to describe the immediate impact of fire on the post-fire environment thereby focusing on short-term changes because of vegetation consumption and mortality, charcoal production and soil alteration [15]

Methods

Results

Discussion

Conclusion