B13E-01
Long term impacts of controlled burns on soil thermal conductivity and soil heating at a Colorado Rocky Mountain site
Heating any soil during a sufficiently intense wild fire or prescribed burn can alter soil irreversibly, resulting in many significant and well known, long term biological, chemical, and hydrological effects. However, much less is known about how fire affects the thermal properties and the long term thermal regime of soils. Such knowledge is important for understanding the nature of the soil's post-fire recovery because plant roots and soil microbes will have to adapt to any changes in the day-to-day thermal regime. Here we report the results of a study performed at a semiarid forest site in the Rocky Mountains of central Colorado (USA) on how fire can affect the long term (post- fire) thermal energy flow in soils. Direct in situ measurements of soil thermal conductivity are presented that show that prescribed burns can alter the thermal conductivity of soils to a depth of at least 0.20 m without altering its bulk density. This is a previously unknown and presently not-well understood effect, which can have significant impacts on the long-term soil thermal regime in itself. Also presented are data on the thermal properties of ash. Such data are necessary for quantifying the impact any remaining post-fire ash layer might have on the daily and seasonal flow of thermal energy through the soil. Additional observational data are presented on the long term effects that prescribed burns can have on soil surface temperatures. Finally, we describe a new analytical model that synthesizes the above observed data to predict the long term changes in the daily and annual cycles of soil heating and cooling that may result from fire. The model incorporates observed (linearly-varying) vertical structure of the soil thermal properties and observed changes in the surface temperatures, to simulate these fire-induced effects. Modeling results suggest that under the dry soil conditions, typical of the experimental forest site, the amplitudes of the daily and seasonal cycles of soil heating/cooling in the fire-affected soils will greatly exceed those in the soils unaffected by fire for several months to years following the fire and that these effects propagate to depths exceeding a meter.
B13E-02 INVITED
Relationships between Fire Radiative Energy, Fuel Combustion Rates, and Emission Rates of Trace Gases and PM2.5 in Laboratory Fires
The MODIS instruments on NASA Terra and Aqua satellites have provided a unique opportunity of measuring fire radiative energy (FRE) with a 1-km x 1-km resolution globally four times every day. The FRE of a fire is affected by meteorology, vegetation types and conditions, and topography. We conducted a series of laboratory experiments to measure simultaneously FRE and emissions of trace gases and PM 2.5 during the combustion of 44 vegetation fires. The dependence of FRE measurements on the instrument viewing angles was also investigated. The vegetation burned included ponderosa pine needles and branches, Douglas-fir twigs and foliage, western white pine needles, sagebrush, and Zambian savanna grass. The results showed that radiative energy accounted for about 12 percent of the total heat released during the fires. The instantaneous fuel weight loss and emissions of CO2, CO, NO and PM2.5 are linearly correlated with the instantaneous FRE release. However, the linear relationships vary considerably for different combustion phases (flaming or smoldering) and fuel types. We also calculated the integrated emission ratios of CO2, CO, NO or PM2.5 to FRE for each experiment. The emission ratios, in combination with the MODIS-derived fire radiative energy, can be used to estimate the amount of trace gases and aerosol particles emitted by fires during the 10-minute satellite overpass time.
B13E-03
Pyromineralization of soil phosphorus in savanna ecosystems
The weathering of rock supplies phosphorus (P) to ecosystems. Phosphorus limitation of ecosystems can be severe in thicker or older soils, where soil production rates from rock and therefore release of P is slower than in thinner or younger soils. Limitation may be especially pronounced in drier ecosystems that are experiencing increasing N deposition. Our savanna field sites in Kruger National Park, South Africa meet all three of these criteria: soil residence times average 250 ky, the climate is semiarid, and N inputs average 20 kg ha-1 y-1. Not all soil P is plant-available, and because our field sites experience occasional fires, our objectives were to quantify the importance of pyromineralization of soil P, the transfer by fire of soil P from recalcitrant to labile (HCO3- extractable) pools. We quantified these soil P pools using a modified Hedley scheme (an array of chemical extractants). Three sets of soils were fractionated: 1. soils from 10 profiles along an intensively studied hillslope, bracketing a pronounced structural and functional ecotone; 2. surface soils from these 10 profiles after a simulated burn; and 3. surface soils from the Shabeni Experimental Plots, where 4 fire treatments have been maintained for decades: no fire, annual fire in the dry season, triennial fire in the dry season, and triennial fire in the wet season. Total P for hillslope soils ranged from 45 to 135 g m-2 (to 50 cm depth) and from 8 to 15 g m-2 (to 5 cm depth). Total soil P was lowest in midslope soils, where upslope sandy soils dominated by broad-leafed vegetation shift abruptly to downslope clayey soils with fine-leafed vegetation. Simulated fire for the hillslope soils reduced total P slightly, but boosted labile P by 1.7 g m-2 (to 5 cm), representing 17% of total P in the surface 5 cm. This pyromineralization effect was not uniform across the hillslope: downslope soils gained about 50% more labile P than midslope soils with simulated burning. With a fire return interval of 4 years, pyromineralization produces a flux of P into plant-available pools of +0.4 g m-2 y-1, a flux that approximates plant uptake of labile P based on foliar P. We observed the same general patterns for soils burnt triennially in the wet season. There, the difference in labile P was greatest between treatment and control plots (10 ug g-1 or 0.9 g m-2 to 5 cm depth), producing a flux of labile P of +0.3 g m-2 y-1. Soils from the other two fire treatments contained less labile P than the control plot, with the most pronounced deficit (-0.25 g m-2 to 5 cm) for the plots burnt triennially in the dry season. Infrequent, high severity fires may prevent the incorporation of biomass P into soils, eliminating the opportunity for pyromineralization of recalcitrant P. These fire-mediated transfers from recalcitrant to labile P pools are nearly 1000 times greater than our estimated flux from rock to soil of total P (0.0008 g P2O5 m-2 y-1). These patterns suggest that pyromineralization represents a significant input of labile P to these P-limited ecosystems, where low chemical weathering rates lead to intense cycling of rock-derived nutrients.
B13E-04
Mapping Fine-Scale Fire Effects After Wildfires in Alaska's Boreal Forest
In recent years, large, severe wildfires have burned in Alaska, which stands to continue as global climate trends continue to be warmer and drier. In 2004, 26,700 km2 (6.6 million acres) burned, and 19,000 km2 (4.7 million acres) in 2005. By these numbers, nearly 10% of Alaska's boreal forest burned in these two years. The boreal forest biome contains a significant percentage of the world's carbon stored as moss and in the highly organic soils. Thus, the extent and severity of fires in the boreal forests is important on a global scale due to the level of carbon emissions from a high amount of organic consumption. To investigate fine-scale fire effects on vegetation and soils, we collected airborne hyperspectral imagery on three interior Alaska wildfires in 2004. The interior area of Alaska is dominated by black spruce forests with dense mats of feather moss on the ground. We applied a five-endmember spectral unmixing model representing green feather moss, a green shrub species, charred feather moss, ash and rock to calibrated reflectance data. The result was fractional cover maps of each input cover type. Significant correlations between the field and remotely sensed data indicated the map was representative of the burned area. The pixel size of the hyperspectral data was 3 m on the ground, meaning the percent of charred and uncharred vegetation and exposed soil or rock was discernable at this scale. The ability to accurately map these biophysical cover fractions, especially fire effects on surface moss and organic soils, may indicate the degree or amount of consumption which can be related to carbon emissions.
B13E-05 INVITED
An Integrated Suite of Multi-scale Remote Sensing Approaches for Measuring Thermal Characteristics and Associated Effects During Wildland Fires
Historically, much of the data used to describe the behavior and characteristics of wildland fires have been derived from post-fire observations, anecdotal evidence, or modeling, rather than from in situ measurements of the fire phenomenon. When used to develop cause-and-effect relationships with respect to first- and second- order fire effects, these inferences frequently become circular. This paper presents a highly integrated fire remote sensing study designed to explore the relative accuracy and efficacy of alternative remote sensing methods for characterizing fires and for predicting first-order fire effects. For this study, we developed new instruments and protocols and also adapted state-of-science technology to acquire remotely-sensed data at multiple temporal and spatial scales on wildland fires. As many as four remote sensing approaches were implemented in a hierarchy of both space and time on each of eight field deployments during four wildland fire incidents. Approaches included in situ kinetic and radiometric thermal observations within a plot area which is to be burned over by fire; oblique-angle thermal radiometry from a viewpoint outside of the instrument deployment site; multi- spectral thermal infrared image acquisition from fixed-wing aircraft; MODIS sensor fire detection and mapping. In addition, both pre- and post-fire ground sampling was performed within the plot to characterize fuel loadings and moistures, vegetation, overstory, and site attributes. In this paper, I present and describe new remote sensing technologies, compare their relative performances through assessment of data acquired at four spatial scales, and we discuss the relationship of the thermal flux data to first-order fire effects at multiple field deployment sites.
B13E-06
Biomass Combustions and Burning Emissions Inferred from GOES Fire Radiative Power
Biomass burning significantly affects air quality and climate changes. Current estimates of burning emissions are rather imprecise and vary markedly with different methodologies. This paper investigates biomass burning consumption and emissions using GOES (Geostationary Operational Environmental Satellites) WF_ABBA (Wildfire Automated Biomass Burning Algorithm) fire product. In doing this, we establish a set of representatives in diurnal patterns of half-hourly GOES Fire Radiative Power (FRP) for various ecosystems. The representative patterns are used to fill the missed and poor observations of half hourly FRP in GOES fire data for individual fire pixels. The simulated FRP is directly applied to the calculation of the biomass combusted during fire activities. The FRP-based biomass combustion is evaluated using the estimates using a traditional model which integrates burned area, fuel loading, and combustion factor. In the traditional model calculation, we derive burned areas from GOES WF_ABBA fire size. Fuel loading includes three different types (1) MODIS Vegetation Property-based Fuel System (MVPFS), (2) National Dangerous Rating Systems (NFDRS), and (3) the Fuel Characteristic Classification System (FCCS). By comparing the biomass combustions across the Contiguous United States (CONUS) from 2003-2005, we conclude that FRP is an effective tool to estimate the biomass burning emissions. Finally, we examine the temporal and spatial patterns in biomass combustions and emissions (PM2.5, CO, NH3) across the CONUS.