Understanding burn severity sensing in Arctic tundra: exploring vegetation indices, suboptimal assessment timing and the impact of increasing pixel size

摘要: Little is known about how satellite imagery can be used to describe burn severity in tundra landscapes. The Anaktuvuk River Fire ARF 2007 burned over 1000 km2 of on the North Slope Alaska, creating a mosaic small 1 m2 large >100 patches that differed severity. scar provided us with an ideal landscape determine if single-date spectral vegetation index once recovery began and independently pixel size influences assessment. We explore sensitivity several commonly indices variation across influence assessment classification conducted field surveys reflectance at peak first growing season post-fire extended period 18 sites ranged from high low In comparing indices, we found two-band enhanced EVI2 was highly correlated normalized ratio NBR better distinguished among three classes than both difference NDVI. also show clear evidence shortwave infrared SWIR reflectivity does not vary as function By Quickbird scene 2.4 m pixels simulated 30 250 m scenes, are able confirm while moderate spatial resolution Landsat Thematic Mapper TM sensor m sufficient for mapping severity, coarser Moderate Resolution Imaging Spectroradiometer MODIS 250 well matched fine scale heterogeneity scar.