Key points are not available for this paper at this time.
Forest depletions caused by fire can be detected by remote sensors if the change event causes a change in surface reflective or thermal properties. Pre and post-fire TM imagery of a semi-arid region was used to map the burn scar of a management-ignited fire into three classes of fire severity. The multitemporal imagery was enhanced using several brightness, greenness, and wetness indices. The wetness indices were most accurate at delineating fire severity because fire severity appears related to changes in plant and soil moisture content. Overall kappa for Kauth Thomas Δ wetness, the TM 7/4 index, and the second standardised Principal Component were 0.62, 0.59, and 0.61 respectively. The highest overall kappa of 0.66 was achieved using combined Kauth Thomas Δ brightness, Δ greenness and Δ wetness indices.
Rogan et al. (Mon,) studied this question.