B41B-0100 0800h
Charcoal Morphology, a Useful Indicator of Fire Signature in Prosser Lake, British Columbia, Canada.
Observations of charcoal particle size and morphology were used along with quantitative analysis to extract indications on fire events over the last century from Prosser Lake (49\deg45.05 N, 120\deg37.30 W), a mesotrophic and slightly meromictic lake from British Columbia, Canada. Charcoal particles $<$150$\mu$m were visually identified at a 0.5-3 years resolution in a laminated sedimentary sequence. According to their shape and structural features, charcoal particles were classified in seven morphologically distinct types and their abundances were assessed using Image Analysis techniques. Distributions of charcoal types were assessed as a proxy to fire events recorded between 1919-2000 and subsequent mechanisms of transportation-sedimentation to lake sediments. Frequent fires taking place before 1944, produced high amounts of charred particles, but strong fires that took place in 1939, 1940, 1958 and 1960 were poorly recorded by most of the charcoal types, whereas post-1944 periods of high precipitation levels without fire events increased their abundance. However, fragile-type fragments displaying high porosity walls showed a strong and significant correlation (R$^{2}$ = 0.7; p = 0.02) with historically recorded forest fire in the proximity of Prosser Lake. Those fragments, according to their shape and structure might originate from high fragmentation of wood burning at high temperatures or by burning of small branches and leaves. Being very fragile, particles of this type would be destroyed by eventual secondary transportation processes and would not occur in levels not related to fire events. The remaining types of charcoal from Prosser Lake sediments displayed distributions biased by secondary transportation-sedimentation processes. We propose that charcoal morphology can be a useful indicator of fire occurrence, proximity of source-area and transportation-sedimentation mechanisms.
B41B-0101 0800h
The Role of Fire History in Northern Latitude Organic Matter: Spatial and Temporal Patterns of C and Hg
Fire history has dictated to a large extent the spatial occurrence and chemical composition of organic matter stored in postglacial peat and upland soils of northern latitudes. These regions are characterized by a unique co-occurrence of discontinuous permafrost and pervasive wetlands with large fires and large stocks of ground fuels. The distribution of combustible elements on the landscape is telltale of fire histories that have varied widely across these landscapes: Low C and Hg stocks in well drained uplands is consistent with modern radiocarbon signals in organic layers above mineral soil. Larger C and Hg stocks in poorly drained uplands are consistent with less frequent and/or less severe fire histories. Also, radiocarbon profiles demonstrate organic matter accumulation over the past 100 - 200 years. C and Hg in very poorly drained bogs and fens are highest of all systems, and high concentrations of Hg within 20 cm of the modern water table suggest that fires have occurred either infrequently or with low severity as compared to upland sites. Barring significant differences in losses to decomposition and leaching, fire emissions of C and Hg also must be dictated by landscape patterns of soil drainage and therefore by climate-drought-water table interactions. We project estimates of Hg emissions from Hg inventories of lowland and upland soils combined with landscape assessments of lowland and upland distributions and find that Hg likely varied by at least 5-fold as a result of temporal variations in drought, burn area, and water table depression. With the increasing threat of Hg toxicity in northern latitudes, questions of paleofire, paleoclimate, and past landscape distributions of fire and organic matter become increasingly important to our need for defining future Hg emissions, deposition, and redistribution on the landscape.
http://carbon.wr.usgs.gov
B41B-0102 0800h
Fire-Climate Linkages in the Northern Rocky Mountains During the Holocene
Foy Lake, located in low elevation forests of northwestern Montana, provides records of annual-to-decadal-scale landscape change for the last 13,000 years. Sedimentary charcoal and pollen analyses were used to document fire-climate linkages in the Northern Rocky Mountains (NRM) to improve the interpretation of lake-sediment records of fire activity. This paper focuses on the climate controls that suppress and promote fire activity on seasonal-to-decadal time-scales and the ability of lake-sediment proxies from the NRM to record past fire activity occurring at different spatial and temporal scales. Annually-sampled charcoal in lake-sediment cores is compared with evidence of known fires identified from historic fire atlases and fire-scarred tree-ring studies. Historical records of fire near Foy Lake, including the AD 1910 fires that burned over two million acres in Idaho and Montana, are used as a baseline for the analysis of sedimentary charcoal influx during Holocene fire episodes. During recent drought and fire years in the NRM, anomalously higher-than-normal 500mb geopotential heights in summer extended across the Pacific and into northwestern North America and produced stronger westerlies, increased subsidence off the eastern north Pacific and western North America, and suppressed precipitation. Large fire years in the early 20th century, including AD 1910, are associated with above average spring precipitation, higher-than-normal spring temperatures, followed by lower-than normal summer precipitation. The long-term record of sedimentary charcoal and pollen at Foy Lake shows that fire activity (inferred from the frequency of charcoal peaks) was initially low during the early Holocene, increased between 7000 to 5000 cal yr BP, and then decreased from 3800 to 2200 cal yr BP. Fire frequency increased from 7 episodes/1000 years at 2200 cal yr BP to 14 episodes/1000 years at 800 cal yr BP. Fire episodes reached a maximum at ca. 500 cal yr BP and then decreased to present levels.
B41B-0103 0800h
Global Fire Patterns and Trends from ENSO Events Estimated Using an Enhanced Ecosystem Model
Fires play a major role in structuring terrestrial ecosystems and in the exchanges between the biosphere and the atmosphere. They can affect vegetation density, size distribution, species composition, carbon and nutrient content of soils, as well as particulate and trace gas emissions. Since fires are strongly influenced by climate, their location, frequency, and severity may also be affected by climate change and variability. Models are needed to estimate the patterns and consequences of future fires. For reliability, these models should be validated with information from the past. A fire sub-model driven by climate and biomass was implemented in the 1°x1° resolution Global Terrestrial Ecosystem Carbon model (GTEC 1.0) and was run for the years 1901 to 1998. Simulated fires were aggregated by vegetation type and by region and compared favorably to patterns from satellite-based fire products. For example, in 1997 model estimates of regional fire activity were within 1.4% in Northern Asia, 4.2% in Central America, 9.6% in North America, 11.5% in Europe, and 16.9% of all regions averaged globally when compared with the European Space Agency Fire Atlas. In order to identify spatial and temporal patterns that specifically relate to the El Nino Southern Oscillation (ENSO) phenomenon, model estimates were compared to the Southern Oscillation Index (SOI). The model correctly predicted fire variability in regions strongly affected by ENSO. Results suggest that this model could be used to predict global fire trends under climate change scenarios, and used as a component in Earth system models to assess the many potential consequences of altered global fire patterns.
B41B-0104 0800h
An Assessment of Ecosystem Effects of Wildfire from Space: Case Study of the 2002 Okefenokee Fire
Wildfire is a major natural disaster threatening ecosystem, human life and property in the United States. Efforts have been made in the past years in monitoring and understanding wildfire-ecosystem interactions. Satellites Remote Sensing (RS) measurements can provide nearly global coverage of the Earth with high spatial resolutions and frequent repetition rates. Satellite products such as the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI) and Leaf Area Index (LAI) from the Advanced Very High Resolution Radiometer (AVHRR) and the Moderate Resolution Imaging Spectroradiometer (MODIS) have shown close relationships between burned areas and ecosystem changes. Thus, satellite RS has emerged as one of the important techniques for wildfire-ecosystem interaction research. MODIS NDVI, EVI and LAI datasets are analyzed. This paper illustrates the impacts of wildfire on ecosystem by applying the MODIS measurements to the 2002 Okefenokee fire. The preliminary results shown EVI and LAI are better than NDVI for assessing Ecosystem Effects of Wildfire in this region. Key words: wildfire, ecosystem, MODIS, and satellite remote sensing.
B41B-0105 0800h
Dependence of Soil Temperature on the Fluxes of CO$_{2}$, CH$_{4}$, and N$_{2}$O in Boreal Forest Soils, Interior Alaska
We measured the fluxes of CO$_{2}$, CH$_{4}$, and N$_{2}$O using static chambers and monitored soil temperatures at three depths (5, 10, and 20 cm) in burned and unburned black spruce forest soils, interior Alaska during the growing seasons of 1999 to 2002. The FROSTFIRE burning experiment was carried out in 1999 to focus on the large-scale ecological consequences of fire. We observed significant dependence of fluxes of CO$_{2}$ and N$_{2}$O on soil temperature at depths of 5, 10, and 20 cm in burned and unburned black spruce forest soils. CH$_{4}$ fluxes also showed a weak exponential correlation with soil temperature at the three depths. The effect of soil temperature at the 5, 10, and 20-cm depths can be explained the maximum of 50%, 40%, and 35% of the variability of annual Q$_{10}$ values for both fluxes, respectively. Therefore, soil temperature appears to be one of the significant factors determining the soil-originated fluxes in boreal forests soils of central Alaska.
B41B-0106 0800h
Terrestrial Productivity in USA derived by A Biogeochemical Model and Remote Sensing Products from 1982 to 1997
In conterminous USA, Biome-BGC uses the 1980-1997 meteorology data, STATSGO soil texture data for sand/silt/clay percentages and effective soil depth, MOD12Q1 land cover type 5, and elevation data from GTOPO30 aggregated to 8 km resolution as the input data. MOD 17 algorithms use remotely sensed AVHRR, FPAR, LAI and DAYMET meteorology data as the input. GPP/NPP mean and NPP std simulated by Biome-BGC and MOD 17 have the similar patterns in conterminous USA, with Biome-BGC results lower in the middle from north to south during 1982-1997. For the annual average over conterminous USA, solar radiation (SRAD) affects GPP/NPP the most, precipitation the second and VPD the third. The relationships between GPP, NPP and mean temperature are not significant. For the spatial results, the annual mean temperature decreases in the middle and east part, increases in most of western part. The precipitation decreases in the western-southern and east part around the five-lake part, increases in the middle, western-northern and east parts of conterminous USA. VPD (vapor pressure deficit) has the similar change as temperature, increases mainly in the west-south part. For the Biome-BGC results, GPP/NPP decrease in the northern-eastern, middle and western-southern parts. For the meteorology variables trends, precipitation conflicts with NPP in the south-west, agrees with GPP/NPP in the middle and east.
B41B-0107 0800h
Modeling the Dynamics of Wildfire in the North American Boreal Forest
In the boreal forest, where carbon stores are large, wildfire is a common occurrence, and changes in the fire regime have consequences for water, energy, and trace gas feedbacks to the climate system. Although fire is common, the underlying mechanisms that influence the fire regime at large scales are incompletely understood. Because changes in fire regime have implications for the climate system, it is important to develop a predictive capability to simulate the response of fire regime to future climate changes. In this study, we developed a fire model with large spatial coverage and compared the simulations of area burned in Alaska and Canada between 1959 and 1995 to observations over that period. This comparison verified that the model accurately represents the fire regime of the North American boreal forest. We coupled the fire model to the terrestrial ecosystem model (TEM) and simulated carbon dynamics of Alaska and Canada for the period 1860-1995. The carbon dynamics simulated for Alaska and Canada were similar to the carbon dynamics of a simulation that was driven by observed fire history from 1959-1995. Our next step in this study is to use the coupled model to evaluate carbon dynamics for future scenarios of climate change.
B41B-0108 0800h
Predicting the Global Distribution of Fire Emissions: Can Supervised Classifiers Inform Physical Models?
An accurate prediction of fire emissions is an important piece of the climate variability puzzle. Accurately predicting emissions depends on an accurate prediction of ignition events, fuel moisture levels, fire spreadrates, and combustion completeness of different fuel types. Current global vegetation models predict fire ignition and spread using simplified models that depend primarily on fuel load and litter moisture. We used data-mining and machine-learning techniques to find relationships between potential drivers of fire and the occurrence of fire (as measured by the TRMM-VIRS satellite). We compared the fire prediction accuracy of physical models and supervised classifiers trained on 1998-2002 data. We finally discuss how relevant drivers can be incorporated into fire models to improve accuracy of predicted fire emissions.
B41B-0109 0800h
Temporal Patterns of Fire in West Kalimantan, Indonesia
Fire is an essential landscape management tool extensively employed in West Kalimantan Indonesia to clear land and prepare agricultural areas. Under typically wet climatic conditions fires are easily controlled and seldom spread into adjacent land cover. However, during droughts induced by strong El Nino events, land management fires threaten vast areas of the landscape threatening endangered species habitat and releasing large amounts of carbon into the atmosphere. This study investigates temporal and spatial variations of fires detected by Moderate-resolution Imaging Spectroradiometer (MODIS) and Along Track Scanning Radiometer between 2000 and 2004 against the MODIS Vegetation Continuous Fields and cultural features manually digitized from Landsat ETM+ scenes. Patterns of fire during phases of the El Niño-La Niña cycle are described and the impacts of fires on orangutan habitat are investigated.
B41B-0110 0800h
Identification of the Interrelationship Between the Interannual Variation in Vegetation Activity and Climate Over Tropics
Clarification of the global terrestrial carbon cycle is essential to appreciate the green house effect. In special, tropical ecosystems have been influenced by the climate variations such as ENSO. Therefore, estimating the correlation between interannual variability in vegetation activities and the climate system is important. Although many studies have shown the relationship between the vegetation activities and ENSO events, they are not enough to evaluate the interannual variability and each response between the vegetation activities and climate parameters in the system. In this study, we used the 10 days composite Normalized Difference Vegetation Index (NDVI) originated from the Advanced Very High Resolution Radiometers (AVHRR) on the NOAA satellites from 1981 to 2000. Firstly, we estimated the variation associated with ENSO period using Singular Spectrum Analysis (SSA), which is useful method to extract a low frequency component from short-term noisy time-series. Secondly, we proposed the multivariable linear system model using the ENSO-related variation in NDVI, Multivariate ENSO Index (MEI) and the atmospheric CO2 growth ratio, and then investigated their mutual responses and feedback system in the model. Although the NDVI time-series over tropics contain many noisy elements due to cloud contamination or sensor degradation, the principal components ENSO-associated cycle explained a certain proportion of variance and the reconstructed fluctuation showed high correlation with Southern Oscillation Index (SOI), MEI, and ENSO-connected climate variations (precipitation, temperature, and incoming surface solar radiation) on a seasonal basis. Our model results showed that the regions correlated well with ENSO activities are inclined to coincide with the interannual vegetation variation. Our conclusion is that the NDVI time-series over tropics, where have been affected by ENSO activities, contain the ENSO-related component and link with the interannual CO2 growing variation. Future of our studies are to improve the NDVI data set to predict future well and to compare ENSO-associated variation in NDVI with NPP or GPP derived from the biochemical model.
B41B-0111 0800h
Effect of Fuel Moisture on Fire Spread: A Sensitivity Analysis
Fuel moisture, live as well as dead is an important parameter for wildfire danger and wildfire spread. In this work we use the fire behavior model FARSITE to analyze the sensitivity of fire spread on fuel moisture. Fire spread is simulated over two wildfire prone areas in California and Georgia representative of the different conditions in the western and eastern USA. The selected locations are actual wildfire sites and thus give us an opportunity to compare the effect of change in fuel moisture over two different surface types as well as evaluate the performance of the model. The results indicate that fuel moisture effects are different over different surface types and fuel moisture estimation using remote sensing must be considered for operational fire danger rating systems.
B41B-0112 0800h
Soil Carbon and Nitrogen Dynamics in Fire-Suppressed, Wildfire-Burned, and Prescribe-Burned Chaparral in the Sierra Nevada Foothills
Chaparral shrublands cover 13 percent of California's land area and are very dynamic, productive, and flammable. While chaparral supports potentially destructive stand-replacing fires that provide opportunity for rapid, large-scale vegetation manipulation for range management and fire hazard reduction, there has been little investigative work describing soil organic matter dynamics in chaparral. We report findings of research on vegetation cover and soil organic matter fractionation in passive, slow, and rapidly cycling pools under different fire history regimes and soil types in chamise-manzanita-toyon chaparral of the Sierra Nevada Foothills. We used detailed fire history overlays and soil-vegetation maps to identify five fire history scenarios: 1) long-term fire-suppressed; 2) 20-year wildfire frequency (1950-1972-1992); 3) four-year wildfire frequency (1997-2001); 4) one-time wildfire (2001); and 5) prescribed reburn (2001-2005-2006) where the management goal is to type convert dense, resprouting shrub cover to a grass-shrub mosaic in a strategic fire hazard reduction zone. We replicated these sampling areas on three soil types ubiquitous to Sierra Nevada Foothills chaparral: soil derived from residuum of granite, basic igneous and metamorphic rocks, and serpentine. Vegetation cover analysis shows that the four-year fire recurrence interval has significantly lower shrub cover and higher annual grass cover compared to all the other treatments. Results of vegetation cover analysis, soil organic carbon fractionation, estimates of microbially available carbon, and indices of potentially mineralizable nitrogen will be presented.
B41B-0113 0800h
The Construction of Scrub in California and the Mediterranean Borderlands: Climatic and Edaphic Climax Mosaic or Anthropogenic Artifact?
There is a marked difference in the representation of Mediterranean scrub vegetation (e.g., chaparral, maquis) in North American and European literature in biogeography and ecology. Authors discussing this vegetation in the California context accept that it is a natural response to the Mediterranean climates, with their late summer and fall fires, and steep terrain. Debate here focusses on the extent to which humans have modified or, indeed, can modify "natural" fire regimes. European authors frame this vegetation instead as a secondary successional formation in a landscape that "should" be dominated by oak woodland and forest. The widespread presence of Mediterranean scrub is cast as an artifact of human disturbance over thousands of years, mediated through overgrazing, deforestation, accelerated erosion, and anthropogenic fire. This poster will present a content analysis of the Mediterranean scrub literature, in order to engage both traditions in the construction of a unified framework for these pyrogenic formations.
B41B-0114 0800h
Long-term Stream Flow Impact Of Wild Fires In Mediterranean Shrubland Ecosystems
In Southern California, chaparral ecosystems dominate the local mountains. These ecosystem have a fire return interval under natural conditions of 40 to 60 years. Fire is an integral process to these ecosystems fostering germination and regular changes in vegetation. Inhabitants of Southern California depend on chaparral catchments for upwards of 1/3 of their water supply through rainfall in the mountains and recharge to regional aquifers at the mountain front. There is ample evidence in the literature to support the idea that soon after a wild fire, runoff increases. However, the long-term effects of burns that cover greater than 50 percent of a watershed have been investigated to a lesser degree. The aim of this paper is to begin preliminary investigation of the long-term effects of large wildfires. The results after examination of two sets of paired watersheds so far show very little impact after 3 to 5 years, with flows returning to a statistically insignificant difference within that time period. Three other watersheds preliminarily chosen for this study showed little impact from fire due to evidence of long term precipitation variability and its impact on annual stream flow. This result clearly shows that not all variables were accounted for in this study and that long term data ($<$ 100 years) may be needed in some situations to demonstrate what the effects of fire are on annual stream flow.
http://soilsa120a.ucr.edu/~escott/WildFire/
B41B-0115 0800h
Streamwater Chemistry and Sediment Responses to Wildfire in the Colorado Front Range
The influence of forest fire on streamwater chemistry depends on the extent and conditions of the burn, the physical and biotic characteristics of the watershed and the flow regime. A monthly streamwater monitoring network initiated in 2001 on the Pike National Forest allows evaluation of fire effects in catchments burned by the 2002 Hayman fire and allows comparison of streams in burned and unburned drainages. In burned watersheds, volume-weighted suspended sediment increased by an order of magnitude during the first post-fire year and by another order of magnitude the following year. Stream nitrate concentrations increased linearly following the burn. The year following the burn, discharge-weighted concentrations of Na$^{+}$, K$^{+}$, Mg$^{+2}$ and Ca$^{+2}$ differed little between streams in burned and unburned watersheds; however, in year 2 these constituents increased by 51, 13, 30, 35 and 51%, respectively for burned drainages. Similarly, differences in streamwater ANC, NO$_{3}$$^{-}$ and Cl$^{-}$ between burned and unburned drainages were more pronounced the second year after the burn. We will also assess the influence of watershed size and fire severity on the responses of individual streams.
B41B-0116 0800h
Estimating Effects of Brazilian Forest Wildfires on the Carbon Monoxide Concentration
Forest wildfires have dramatically increased in recent years due to global warming and extreme dry conditions. Forest wildfires spew out a significant amount of atmospheric pollutants, such as carbon monoxide, due to incomplete burning of the biomass. According to United State Environmental Protection Agency (EPA), a high increase of carbon monoxide leads to the formation of carboxyhemoglobin in blood which decreases the oxygen intake capacity of human body and at moderate concentration angina, impaired vision and reduced brain function may occur. As compared to Northern America where significant amount of carbon monoxide released is caused by combustion devices and furnace, the increase of carbon monoxide concentration in Brazilian regions is mainly attributed to the forest fires. In this study, carbon monoxide datasets from the Measurements of pollution in the troposphere (MOPITT) have been analyzed to see the amount of increase in the carbon monoxide concentration after forest wildfires, ire, particularly in summer of 2003. The study reveals that there is a significant increase in the carbon monoxide concentration after forest fires.
B41B-0117 0800h
Interannual Variability of Global Biomass Burning Using Atmospheric Measurements of Carbon Monoxide and Hydrogen
Biomass burning is a significant contributor to the global budgets of many environmentally important trace gases. Its sources are both natural and anthropogenic and any number of variables can affect its emissions over annual and decadal timescales. An important outstanding question in global change science is, have the emissions of biomass burning changed appreciably over recent years? We use a novel technique to address this question utilizing global atmospheric measurements of carbon monoxide (CO) and hydrogen (H2). These are both produced in large quantities by biomass burning, but are removed by different processes. We invert the measurements using a two-dimensional transport model to obtain a twenty-year time series of their respective sources. We find that these source histories are in good agreement with one another and share similar features and interannual variations. As the sinks for these gases are different, the agreement must reflect changes in a common source. If we restrict our analysis to the tropics where biomass burning is greatest, the main common sources of CO and H2 are oxidation from methane and non-methane hydrocarbons (NMHCs), and biomass burning. We argue that variability in the source time series most likely reflects changes in biomass burning, as we can account for the methane oxidation source, and the other source is dominated by natural NMHCs, which are not expected to vary significantly. We conclude that though the biomass burning record has large interannual variability, there is no evidence for a significant decadal trend. This work was supported in part by a grant from the Department of Energy (Grant DE-FG03-97ER62401) and the resources of the Andarz Co.
B41B-0118 0800h
Characterizing Wildfire Regimes and Risk in the USA
Over the last decade, high profile wildfires have resulted in numerous fatalities and loss of infrastructure. Wildfires also have a significant impact on climate and ecosystems, with recent authors emphasizing the need for regional-level examinations of wildfire-regime dynamics and change, and the factors driving them. With implications for hazard management, climate studies, and ecosystem research, there is therefore significant interest in appropriate analysis of historical wildfire databases. Insightful studies using wildfire database statistics exist, but are often hampered by the low spatial and/or temporal resolution of their datasets. In this paper, we use a high-resolution dataset consisting of 88,855 USFS wildfires over the time period 1970--2000, and consider wildfire occurrence across the conterminous USA as a function of ecoregion (land units classified by climate, vegetation, and topography), ignition source (anthropogenic vs. lightning), and decade (1970--1979, 1980--1989, 1990--1999). We find that for the conterminous USA (a) wildfires exhibit robust frequency-area power-law behavior in 17 different ecoregions, (b) normalized power-law exponents may be used to compare the scaling of wildfire burned areas between regions, (c) power-law exponents change systematically from east to west, (d) wildfires in 75% of the conterminous USA (particularly the east) have higher power-law exponents for anthropogenic vs.\ lightning ignition sources, and (e) recurrence intervals for wildfires of a given burned area or larger for each ecoregion can be assessed, allowing for the classification of wildfire regimes for probabilistic hazard estimation in the same vein as is now used for earthquakes. By examining wildfire statistics in a spatially and temporally explicit manner, we are able to present resultant wildfire regime summary statistics and conclusions, along with a probabilistic hazard assessment of wildfire risk at the ecoregion division level across the conterminous USA.
B41B-0119 0800h
An update on The Hazard Mapping System (HMS) - a multiplatform remote sensing approach to fire and smoke detection
The Hazard Mapping System (HMS) is a multiplatform remote sensing approach to detecting fires and smoke over the US, Canada, Mexico and Central America. This system is an integral part of the Satellite Services Division's near realtime hazards detection and mitigation efforts. The system utilizes NOAA's Geostationary Operational Environmental Satellites (GOES), Polar Operational Environmental Satellites (POES), the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument on NASA's Terra and Aqua spacecraft and the Defense Meteorological Satellite Program Operational Linescan System (OLS) sensor, (F14 and F15). Automated detection algorithms are employed for each of the satellites (except DMSP OLS) for the fire detects while smoke is annotated by a satellite analyst. Fire detects can also be added by the satellite analyst. Major customers for this product include the National Weather Service, Storm Prediction Center, US Forest Service, Environmental Protection Agency, research science teams, as well as numerous federal, state and local land and air quality managers. In 2004 the HMS was upgraded by adding the Canadian, Mexican, and Central American sectors for hotspot and smoke detection. These sectors can be easily turned off or on by changing flags in the system configuration file. This enables analysis in sectors only during their respective burning seasons. The Alaskan and Canadian sectors are typically turned off in the winter season and the Mexican sector is cut off after the March-May burning season. But sectors can also be easily added or restarted if, for instance, smoke from a region is affecting the United States. Various ancillary data sources are used in the HMS to aid the analysis. Stable Lights is a static product that identifies stable sources of light from the OLS sensor and is usually associated with cities and urban areas. It appears on the screen as a transparent overlay on the satellite imagery being displayed. This capability can assist the analyst by screening out heat sources where stable lights are present. Vegetation type and water overlays aid in the decision to add or delete a fire point. Water sources many times may cause false detects due to low sun zenith angles during sunrise and sunset or due to temperature contrast between land and water at night. This overlay aides in quickly finding these false detects. Vegetation overlays enable the analyst by showing what type of land is present near the hotspot in question. For example, fires are more likely in forest or grassland than desert or barren lands. The SSD fire team is currently assessing the feasibility of a descriptive smoke text product and would like to incorporate additional datasets for the monitoring of fires, smoke, dust, and air pollution. The HMS is a dynamic product that changes with the needs of our analysts and customers.
http://www.ssd.noaa.gov/PS/FIRE
B41B-0120 0800h
Heat Transfer Processes Linking Fire Behavior and Tree Mortality
Traditional methods for predicting post-fire tree mortality employ statistical models which neglect the processes linking fire behavior to physiological mortality mechanisms. Here we present a physical process approach which predicts tree mortality by linking fireline intensity with lateral (vascular cambium) and apical (vegetative bud) meristem necrosis. We use a linefire plume model with independently validated conduction and lumped capacitance heat transfer analyses to predict lethal meristem temperatures in tree stems, branches, and buds. These models show that meristem necrosis in large diameter (Bi $\geq$ 0.3) stems/branches is governed by meristem height, bark thickness, and bark water content, while meristem necrosis in small diameter (Bi $<$ 0.3) branches/buds is governed by meristem height, branch/bud size, branch/bud water content, and foliage architecture. To investigate effects of interspecfic variation in these properties, we compare model results for {\it Picea glauca} (Moench) Voss and {\it Pinus contorta} Loudon var. latifolia Engelm. at fireline intensities from 50 to 3000 kWm$^{-1}$. Parameters are obtained from allometric models which relate stem/branch diameter to bark thickness and height, as well as bark and bud water content data collected in the southern Canadian Rocky Mountains. Variation in foliage architecture is quantified using forced convection heat transfer coefficients measured in a laminar flow wind tunnel at Re from 100 to 2000, typical for branches/buds in a linefire plume. Results indicate that in unfoliated stems/branches, {\it P. glauca} meristems are more protected due to thicker bark, whereas in foliated branches/buds, {\it P. contorta} meristems are more protected due to larger bud size and foliage architecture.
http://www.bio.ucalgary.ca/divisions/ecology/johnson.html
B41B-0121 0800h
Short-term increases in Summer Temperature and Rainfall do not Lead to Carbon Storage in the High Arctic
From landscape to global scales climatic controls over primary production and ecosystem respiration are evident through the pattern of carbon accumulation in vegetation and soils. Few ecosystems, however, may be as constrained by climate as those within polar desert landscapes where ecosystem carbon accumulation is low, the average annual temperature is well below zero and precipitation is often less than 15 cm. To characterize how climate affects ecosystem carbon exchange in this cold, dry environment, we have established two experiments: 1) a factorial increase in summer temperature and rainfall and 2) a two-level warming experiment that includes ambient, and 2 and 4 degree C warming treatments. Within these experiments, we have measured net ecosystem carbon exchange (NEE), gross primary production (GPP), and ecosystem respiration (Re) during an extremely dry summer and a summer with average rainfall. Overall, we found that the ecosystem was a net source of carbon to the atmosphere in both years with an increase in carbon loss in the wetter year due to a greater increase in Re than GPP. An experimental increase in rainfall and higher yearly rainfall both led to greater GPP and Re. Increased temperatures only led to higher GPP in the year with more rainfall, and this response was non-linear. Warming increased respiration rates in both years. Although we can observe that carbon has accumulated in this landscape through time, it is not clear what climatic conditions lead to a net carbon gain in the High Arctic.