A52C-05
The influence of synoptic-scale climate variability on boreal forest fire emissions and trace gas variability in Alaska
Friday, 18 December 2015: 11:20
3006 (Moscone West)
Elizabeth Brooke Wiggins1, Sander Veraverbeke1, John Henderson2, Anna Karion3, Jakob Lindaas4, John B Miller3, Roisin Commane4, Steven C Wofsy4, Charles E Miller5, James Tremper Randerson1 and Colm Sweeney3, (1)University of California Irvine, Department of Earth System Science, Irvine, CA, United States, (2)Atmospheric and Environmental Research, Lexington, MA, United States, (3)NOAA Boulder, ESRL, Boulder, CO, United States, (4)Harvard University, Cambridge, MA, United States, (5)NASA Jet Propulsion Laboratory, Pasadena, CA, United States
Abstract:
We examined climate controls on fire emissions and trace gas variability using three distinct conceptual models of fire emissions that draw upon different types of remote sensing information. The three approaches were derived from satellite-derived observations of active fires, satellite-derived estimates of fire radiative power, and daily emissions estimates from the Alaska Fire Emissions Database model (AKFED). In our analysis, we assessed the relative importance of different climate variables and fire weather indices in explaining the temporal variability of satellite-detected fire thermal anomalies and emissions within the state of Alaska during the summer of 2013. We evaluated the performance of each emissions model using trace gas observations from the CARVE (CRV) tower in Fox, Alaska. In our approach we used an inverse atmospheric transport model, the coupled Weather Research and Forecasting/Stochastic Time-Inverted Lagrangian Transport (WRF-STILT) model, to link the fire emissions with the trace gas observations. MISR plume observations were used to inform the injection height distribution in WRF-STILT and CRV-derived estimates of CO/CO2 emission ratios were used to convert modeled carbon emissions into trace gas fluxes. Local climate variables had varying levels of influence on fire dynamics in interior Alaska, with vapor pressure deficit explaining more variability than temperature or relative humidity alone. Combined use of the emissions products and WRF-STILT allowed us identify fire contributions to the CRV time series on a daily basis, and to isolate contributions from individual fires that had different temporal dynamics and interactions with atmospheric transport. Both hourly and daily observed CO mole fractions were highly correlated with the emissions models convolved with WRF-STILT (r2 values up to 0.63), providing some confidence in the ability of satellite-constrained models to capture daily and synoptic patterns in fire emissions.
