GC026-0016
INITIAL WILDFIRE CARBON LOSS QUANTITIES FOR ABOVE AND BELOW-GROUND SOURCES IN WATERTON LAKES NATIONAL PARK, CANADA

Tuesday, 8 December 2020
Poster
Sam Gerrand1, Laura Chasmer1, Jesse Aspinall2, Chris Hopkinson1 and Thomas Jensen3, (1)University of Lethbridge, Lethbridge, AB, Canada, (2)University of Lethbridge, Geography, Lethbridge, AB, Canada, (3)University of Lethbridge, Geography and Environment, Lethbridge, Canada
Abstract:
Although North American forests are carbon (C) dense areas, direct C emissions from wildfires have been difficult to quantify due to uncertainties in the proportion of C lost from above-ground forest biomass compared to combusted organic soils. While C loss from soils is rarely quantified, soils are important sources of C in productive mountainous environments, accounting for ~50% of total ecosystem C.

In 2017, the Kenow wildfire burned 19 303ha (38%) of Waterton Lakes National Park; the study site is within a severely burnt riparian region of the Cameron Lake Valley, approximately 500m north of Cameron Lake. We use field and lidar data to (a) determine the spatial distribution of depth of soil burn within a wet spruce dominated riparian zone and a dry lodgepole pine dominated upland zone ; and (b) determine the variable proportion of C loss from aboveground biomass and organic soils to identify spatial distribution in aboveground and belowground C losses from wildfire. Using the adventitious root method, we measured burn depth based on standing trees (n=313), while C content of soils was determined from 15 depth stratified soil samples from a nearby unburned forest of the same age, species and similar hydro-climatology. Tree height and diameter at breast height were obtained within six (radius = 11.3m) tree plots in the burned area, and allometric equations were used to calculate aboveground C loss. Airborne multispectral lidar data with three spectral bands (532nm, 1064nm, 1550nm) were collected in July, 2018 and used to discern the spatial distribution of environmental characteristics that may have increased depth of burn. Differential Global Positioning System data from the adventitious root measurements were interpolated to produce a 1m resolution burn depth map. Results indicate an average burn depth of 0.16m (stdev. = 0.13m) with a maximum value of 1.05m proximal to a perennial stream, which may have eroded significantly following a wet spring and lack of organic soils. Current field activities suggest that a non-linear model will best represent exponential C losses in areas of deeper burns—where deeper, compacted peat was available to smoldering combustion due to extended dry conditions. Quantifying sources of C loss are important measurements for improved emissions modelling and impacts on local community health.