Identification of previously unrecognized tundra fire events on the Arctic Slope of Alaska

Tundra vegetation, peat, and frozen soils in the Arctic store large, globally significant amounts of labile carbon and pulse disturbances, such as fire, play an important role in releasing this carbon. Tundra fires on the Arctic Slope of Alaska, however, are reportedly rare over the period of historical record keeping as well as throughout the Holocene. Between 1950 and 2011, 26 fires have been identified and cumulatively these burned a land area of ~150,000 ha. However, the Anaktuvuk River Fire of 2007, accounts for two-thirds of the historical burn area in this region and has prompted the notion that the northern Alaska tundra fire regime may be shifting. Further, based on analysis of charcoal found in lake sediments within the burn perimeter it has been labeled as an unprecedented event, at least over the last 5,000 years. However, the historic and pre-historic context of tundra fires on the Arctic Slope of Alaska remains poorly resolved due to the short duration of record keeping, the remoteness of the region, and the cryptic nature of charcoal dispersal and preservation in the paleoecological record. Thus, is it possible that other large and potentially severe tundra fires occurred prior to 1950? To understand how post-fire vegetation succession and land surface changes could be used to potentially identify unrecognized tundra fire events, we conducted studies on upland tussock tundra presumed to have not burned and similar upland-type settings in a chronosequence of past known tundra fires from 2007, 1993, and 1977. Plot-level vegetation and polygonal ice wedge network microtopography surveys were conducted in each area. We then used remotely sensed data and aerial surveys in an attempt to identify unrecognized burn areas. This coupled analysis resulted in the identification and delineation of three tundra fire burn areas on the Arctic Slope of Alaska that resulted from fires that occurred before 1950. Two of these fires were potentially very large, with one tentatively exceeding 100,000 ha and the other exceeding 50,000 ha, while the third fire only burned a small area (300 ha), but was likely of high burn severity. We will present initial results of our field surveys, burn area perimeters, approximate age of the tundra fires based on radiocarbon dating of charred material preserved in the soil, and implications of these previously unrecognized tundra fire events. In addition to the recently identified pre-1950 tundra fires, we also identified eight small fires during imagery analysis that occurred post-1950 that are not captured in the historical fire database. Thus, it is likely that other such tundra fires, both pre- and post-1950, have occurred on the Arctic Slope of Alaska that remain undocumented. Based on our initial results, the role and frequency of tundra fire as a disturbance to tundra ecosystems in northern Alaska may need to be reassessed.