A comparison of red, NIR, and NDVI for monitoring temporal burn signature change in tallgrass prairie

Abstract

Prescribed burning in tallgrass prairie benefits both human and natural systems. However, negative aspects of burning, such as air pollution, also exist. Balancing the advantages and disadvantages of burning requires knowledge of the burn regime in the tallgrass system. One way to acquire this knowledge is by mapping burned areas with remotely sensed data. Unfortunately, burned area mapping is often complicated by the transient nature of burn scars, by cloud cover, and by a lack of spectral contrast between burned areas and non-burned senescent vegetation. In this study, we used in situ measurements of spectral reflectance to track the efficacy of bandpasses that simulated Moderate Resolution Imaging Spectroradiometer (MODIS) and Thematic Mapper/Enhanced Thematic Mapper Plus (TM/ETMþ) sensor responses. Our results showed that red and Normalized Difference Vegetation Index (NDVI) outperformed Near-Infrared (NIR) for delineating burned areas. From these findings it is clear that identifying/mapping burned areas can be performed accurately for several months with either red reflectance or NDVI.

Post-Fire Effects in Wetland Environments: Landscape Assessment of Plant Coverage and Soil Recovery in the Paraná River Delta Marshes, Argentina

“…At regional to global scales, the detection of burned areas using satellite data has been traditionally carried out by the Advanced Very High Resolution Radiometer (AVHRR) because of its high temporal resolution (Kaufman et al 1990, Chuvieco et al 2008. However, the focus of AVHRR-fire studies to date have been on terrestrial ecosystems like forests (Briz et al 2003, Falkowski et al 2005, grasslands (Kauffman et al 1994(Kauffman et al , 1998Mohler and Goodin 2010), or shrublands (Lozano et al 2007). Few works are available on wetlands (Ramsey et al 2002, Cassidy 2007, probably due to their reduced area, natural fragmentation, and difficulty of access.…”

Identifying a suitable combination of classification technique and bandwidth(s) for burned area mapping in tallgrass prairie with MODIS imagery

“…The red spectral region also performed well according to Lopez-Garcia and Caselles (1991) in forest, and according to Stroppiana et al (2002) in savannah. Both NIR and red were useful for differentiating burned from unburned areas in tallgrass prairie according to Mohler and Goodin (2010). Finally, long wave near infrared (LNIR) was found useful for differentiating burned and unburned areas by Li et al (2004) in forest and by Trigg and Flasse (2000) in savannah.…”

“…These studies stand in contrast to tallgrass prairies, where burning is typically done during the spring growing season. Consequently, a rapid decrease in the char signal, along with rapidly regrowing vegetation, quickly eliminates the spectral differences on which burned area detection depends (Mohler and Goodin, 2010;Trigg and Flasse, 2000). This has the obvious effect of causing confusion between unburned vegetation and older burned areas (Eva and Lambin, 1998).…”

Detection of burned areas with extreme gradient boosting algorithm in a multi-temporal setting (MBAGB)

“…The rest of the indexes are also very sensitive to the changes in live green vegetation or moisture content: The normalized burn ratio 2 (NBR2) [37,38], the burn-sensitive vegetation index (MVI) [39], the mid infrared bispectral index (MIRBI) [40,41]. We also use the near infrared (NIR) band 2 of MODIS, commonly used for monitoring temporal burn signatures [42,43]. Higher values of these indexes are associated with high severity, and negative values are associated with regrowth after the fire.…”