College of Behavioral & Social Sciences
Permanent URI for this communityhttp://hdl.handle.net/1903/8
The collections in this community comprise faculty research works, as well as graduate theses and dissertations..
Browse
2 results
Search Results
Item Evaluating characterization of fire extent and fire spread in boreal and tundra fires of Alaska from coarse and moderate resolution MODIS and VIIRS data(2017-04-04) Loboda, Tatiana; O'Neal, Kelley; Yang, QiSatellite observations of fire occurrence, extent, and spread have become a routine source of information for fire scientists and managers worldwide. In remote regions of arctic and boreal zones, satellite observations frequently represent the primary and at times the only source of information about fire occurrence. While a large suite of observations have been shown to provide beneficial and important information about fire occurrence, coarse and moderate resolution data from polar orbiting satellites in optical and thermal ranges of the electromagnetic spectrum provide the most widely-used observations that characterize on-going burning processes and consistent estimates of fire-affected areas. The reliance of the global community on active fire detections and burned area estimates delivered from the Moderate Resolution Imaging Spectroradiometer (MODIS) raises concerns about the continuity of the data record beyond the lifetime of this mission. The Visible Infrared Imaging Radiometer Suite (VIIRS) operated by National Oceanic and Atmospheric Administration (NOAA) represents the future of satellite fire monitoring within US-designed and operated missions. While some advancements have been introduced into the VIIRS fire detection capabilities, including enhanced spatial resolution of spectral bands aimed at active fire detection, the reduced number of orbital overpasses (only one VIIRS instrument is currently in orbit compared to two MODIS instruments) and other differences in data acquisition open the potential for substantial differences in future fire monitoring and mapping capacity and long-term record compatibility between MODIS and VIIRS observations. This study aims to assess and quantify the differences in characterization of on-going burning processes (including in time of detection, spatial fidelity and extent of fire detection coverage, fire spread rate, and fire radiative power) and post-fire extent within fire events (i.e. burned area mapping) in boreal forests and tundra regions of North America delivered by the MODIS Terra and Aqua collection 6 and VIIRS 750m and 375m active fire products and derived burned area maps. Since VIIRS standard data suite does not include burned area estimates, we used VIIRS and MODIS collection 6 surface reflectance products to generate an annual burned area record using the Regionally Adapted Burned Area algorithm developed specifically for high northern latitudes. Our initial results indicate that despite higher spatial resolution of VIIRS observations, the MODIS record (even from a single satellite) delivers a more comprehensive coverage of on-going burning within the large fire events of the 2014 fire season in the Northwest Territories, Canada. However, while substantial differences in fire characterization exist between the satellite data, there is strong potential for calibration of the data records (particularly for the burned area and fire radiative power estimates) for the two instruments necessary to achieve a consistent long-term record of fire occurrence in the high northern latitudes that would support long-term scientific studies and management decision-making processes.Item Carbon Sequestration and Agents of Woody Encroachment in Southeastern Arizona Semi-arid Grasslands(2014) O'Neal, Kelley; Justice, Christopher; Geography; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Woody encroachment and proliferation within dryland ecosystems is potentially the second largest portion of the North American carbon sink and one of the largest areas of uncertainty. This dissertation examines a semi-arid grassland located in southeastern Arizona to better understand woody encroachment, agents of change, and the resultant carbon storage from 1984-2008. The objectives were to quantify changes in woody cover, rank agent importance, estimate carbon density, and calculate voluntary market value. The first objective of mapping changes in woody cover was addressed using a Landsat time-series to measure woody cover and calculate the change, rate of change, and change relative to initial cover over the 25-year time period. Results show the change in woody cover varies spatially and ranges from approximately -2 to 11% with most areas experiencing a 5% increase and 92% relative increase over initial cover, indicating woody cover nearly doubled in the region. The second objective of ranking the importance of agents was achieved using an ensemble classifier. Agents examined included grazing, number of times burned, soil texture, soil productivity, elevation, slope, aspect, and initial woody cover. Initial woody cover, number of times burned, elevation, and grazing were ranked as the most important agents of woody encroachment, indicating the importance of historical land management and disturbance, frequent fire, topography and correlated precipitation, and land use. The third objective of producing carbon estimates and calculating economic opportunity in the voluntary carbon markets was accomplished by applying cover to biomass, root:shoot, and carbon equations to the final woody plant cover maps to calculate carbon stocks, carbon density, and voluntary market value. Results show very low carbon density in the study area relative to similar ecosystems and other ecosystems in general. Given the insignificant annual accumulation of carbon on the small ownership parcels, current low carbon trading prices, and high beef prices, management for storage is not economically viable in the study area at this time.