gisciencestudent

[gisciencestudent]

Annotated Bibliography of Relevant Sources

Burgan, R, van Wagtendonk, J, Keane, Robert E, Burgan, Robert, & van Wagtendonk, Jan. (2001). Mapping wildland fuels for fire management across multiple scales: Integrating remote sensing, GIS, and biophysical modeling. International Journal of Wildland Fire, 10(3-4), 301–319. https://doi.org/10.1071/WF01028

This study compares the advantages and disadvantages of four different methods of fuel mapping: field reconnaissance, direct remote sensing, indirect remote sensing and biophysical modeling. None of these methods appear to be highly accurate or consistent in and of themselves. Yet, all of these methods represent a means of collecting data for synthesis within a geographic information system for the purpose of creating a map that is useful to land management. The authors propose a strategy which involves classifying biophysical setting, species composition and stand structure to assign fuel models, but acknowledge the enduring need for sensor technology that can penetrate a forest canopy to effectively analyze complex surface fuels.

Church, Richard, Adams, Benjamin, Bassett, Danielle, & Brachman, Micah L. (2019). Wayfinding during a wildfire evacuation. Disaster Prevention and Management., ahead-of-print(ahead-of-print). https://doi.org/10.1108/DPM-07-2019-0216

This will be an interesting article to read when it is published, because it is an example of using GIS to synthesize and analyze empirical data from a wildfire evacuation for the purpose of helping emergency managers to develop more effective wildfire evacuation plans. The authors used network analysis to compare volunteers’ selected routes with the shortest distance routes available, and found that only 31 percent of evacuees took a shortest distance route, and that factors such as the elevation of exits and downhill slope could have impacted wayfinding processes. Although this study is more of a spatiotemporal snapshot, more generalizable results could be produced with additional research.

Coops, Nicholas C., Ferster, Colin J, & Coops, Nicholas C. (2014). Assessing the quality of forest fuel loading data collected using public participation methods and smartphones. International Journal of Wildland Fire, 23(4), 585–590. https://doi.org/10.1071/WF13173

This study looks at the potential for citizens to contribute forest structure and fuels input data for use in the applied geographic information systems used by land managers. Citizen contributions could be especially helpful in data collection over broad areas because accurate characterization of forest fuels is dependent on frequent field measurements, as fuels are spatially variable, can change rapidly due to changing conditions, and are difficult to sense remotely under dense canopy. Eighteen volunteers were recruited at the University of British Colombia, and nine of those had extensive working experience in either wildfire suppression or fuels management. The volunteers used an app on their smartphones to collect and report data. For most components, professional measurements were only slightly closer to reference measurements than volunteered measurements, however, non-professional participants notably overestimated aspect and slope. Overall, when appropriate training is provided and adequate controls for accuracy are incorporated, this study found volunteer data collection to be suitable to help inform forest management decisions.

Danzer, SR, Watts, JM, Stone, S, Yool, SR, Miller, Jay D, Danzer, Shelley R, … Yool, Stephen R. (2003). Cluster analysis of structural stage classes to map wildland fuels in a Madrean ecosystem. Journal of Environmental Management., 68(3), 239–252. https://doi.org/10.1016/S0301-4797(03)00062-8

The authors of this study highlight the importance of quality baseline fuels data since we are not yet capable of assessing understory fuels with remotely sensed data. This assessment therefore combines field data collections with GIS, remote sensing, and hierarchical clustering to map the variability of fuels within and across vegetation types of 156 plots from a mountain range in the southwestern U.S. Vegetation classification was validated using an independent sample of 479 randomly located points and demonstrated substantial accuracy with a Kappa value of .80. However, the overall map, created by combining the land cover/vegetation type classification and fuel classes within vegetation type classifications received a relatively low Kappa of .50. This reduction in accuracy could be attributed to GPS errors, ecological overlap between adjacent vegetation types, and/or confusion of fuel classes in areas where overstory canopies obscured the understory.

Dean, DJ, Blackard, Jock A, & Dean, Denis J. (1999). Comparative accuracies of artificial neural networks and discriminant analysis in predicting forest cover types from cartographic variables. Computers and Electronics in Agriculture., 24(3), 131–151. https://doi.org/10.1016/S0168-1699(99)00046-0

The authors compare an artificial neural networks (ANNs) approach to GIS with a conventional model based on discriminant analysis (DA) with regards to effectiveness at predicting forest cover types. In this study, elevation of each 30x30-m raster cell was obtained directly from USGS digital elevation model (DEM) data. Results demonstrated that both ANN and DA models tended to confuse ponderosa pine, Douglas-fir, and cottonwood/willow cover types with each other, potentially due to geographic proximity. However, overall, the ANN model, with a predictive accuracy of 71.1 percent, was shown to be superior to the DA model, with a predictive accuracy of 58.4 percent.

Eva, E. K. (2010). A method for mapping fire hazard and risk across multiple scales and its application in fire management. Ecological Modelling., 221(1), 2–18. https://doi.org/info:doi/

This study presents an effective approach for mapping fire risk across large, complex geographies containing diverse ecosystems on multiple scales. The author uses FIREHARM, which is a C++ program capable of computing changes in fire characteristics over time using climate data, to predict fuel moisture and corresponding fire behavior, danger, and effects. This model does not provide spatially explicit information concerning fire spread. Instead, it assumes that every pixel or polygon experiences a head fire and simulates the resulting fire characteristics based on ensuing weather factors. A landscape is represented by series of polygons. Each polygon defines an area of similar characteristics (vegetation, fuel, site conditions), is assigned attributes related to fire behavior, and is also assigned a tree list (with attributes of species, diameter, height) which combine to estimate tree mortality for a region. In 2004, FIREHARM was validated by the results of a comparison with 54 sample plots from the Cooney Ridge and Mineral Primm wildfires, producing adequate predictions of fuel consumption within approximately 14 days. The model also had a 60 percent chance of accuracy in predicting canopy vs. non-canopy fire and scorch height and fire severity predictions compared well with observed conditions.

Gatzojannis, S, Galatsidas, S, Kalabokidis, Kostas D, Gatzojannis, Stylianos, & Galatsidas, Spyros. (2002). Introducing wildfire into forest management planning: towards a conceptual approach. Forest Ecology and Management, 158(1-3), 41–50. https://doi.org/10.1016/S0378-1127(00)00715-5

This study explains the process of using a geographic information system to synthesis existing information for the purpose of calculating fire danger and fire resistance per unit area (1 km2) and to map distribution within a forest. First, data is collected, next data is grouped into thematic layers, and finally the layers are synthesized for evaluation. The input data required to produce such a map involves taking an inventory of factors with both horizontal and vertical spatial distribution. Homogeneous information layers, such as landscape, are broken down into a series of variables (vegetation zones, land cover structure, aspect, slope, altitude). External factors include climate, landscape, human impact, and other special factors. Internal factor information layers relate to forest stand structure and describe fire resistance at the forest floor, understory, low crown, middle crown, and high crown levels. The mapping of these factors allows for spatial delineation of fire danger zones which decision makers in determining operational objectives and priorities for large geographical areas.

Ilavajhala, Shriram, Wong, Min Minnie, Justice, Christopher O., Davies, DK, Ilavajhala, S, Min Minnie Wong, & Justice, CO. (2009). Fire Information for Resource Management System: Archiving and Distributing MODIS Active Fire Data. IEEE Transactions on Geoscience and Remote Sensing a Publication of the IEEE Geoscience and Remote Sensing Society., 47(1), 72–79. https://doi.org/10.1109/TGRS.2008.2002076

This paper describes the ways in which the combined technologies of remote sensing and GIS are able to deliver Moderate-resolution Imaging Spectroradiometer (MODIS) active fire data to resource managers and even e-mail customized alerts to users. When used as a mobile service, Fire Information for Resource Management System (FIRMS) is an application that can deliver fire information to field staff regarding potential danger. For example, in South Africa, when a fire is detected either using data from MODIS or from a weather satellite, a text message is sent to relevant personnel who can decide if/what action may be required. This strategy makes satellite-derived FIRMS data more accessible to natural resource managers, scientists, and policy makers who use the data for monitoring purposes and for strategic planning.

Kalabokidis, K. (2013). Virtual Fire: A web-based GIS platform for forest fire control. Ecological Informatics., 16, 62–69. https://doi.org/info:doi/

This research project, supported by the University of Athens in Greece and funded by Microsoft Research, describes the high-tech but user-friendly Virtual Fire system, a web-based GIS platform, which allows firefighting forces to share and utilize real-time data (provided by GPS, satellite, camera) for the purpose of locating resources (vehicles, aircrafts, water tanks) and associated shortest routes, monitoring fire ignition probability, and identifying high risk areas. Data from automatic weather stations also aids in fire prevention and early warning. With the ability to conveniently access this information in synthesized form, managers can design more effective and efficient operational plans. Future considerations involve moving to a cloud-based platform, which would allow for expansion to a broader area and the increased incorporation of mobile devices.

Karlsson Martin, Oskar, Galiana Martin, Luis, Montiel Molina, Cristina, Karlsson Martín, Oskar, & Galiana Martín, Luis. (2019). Regional fire scenarios in Spain: Linking landscape dynamics and fire regime for wildfire risk management. Journal of Environmental Management., 233, 427–439. https://doi.org/10.1016/j.jenvman.2018.12.066

The authors of this study apply socioecological systems theory to the wildfire generations model, which describes and explains the appearance and transformation of large wildfires in relation to landscape dynamics within Mediterranean climatic regions. There is a focus on acknowledging the ways in which humans directly and indirectly affect fire regimes. National forest inventory data and existing maps are the data used to create fire scenarios for the Central Mountain Range in Spain via ArcGIS and SPSS23. Land use and land cover features, which relate certain fuel structures to certain fire behaviors, are assigned to 91 discrete geographical units. The resulting visual comparisons can be used to help managers optimize prevention and suppression strategies.

Koukoulas, Sotirios, Kazanis, Dimitrios, & Arianoutsou, Margarita. (2011). Evaluating Post-Fire Forest Resilience Using GIS and Multi-Criteria Analysis: An Example from Cape Sounion National Park, Greece. Environmental Management., 47(3), 384–397. https://doi.org/10.1007/s00267-011-9614-7

The ability to assess an ecosystem’s resilience, or its capacity to endure disturbances without a state change, is becoming more important in the face of an accelerated decrease in biodiversity and with the projected effects of climate change. This study uses geographic information systems (GIS) to assess post-fire resilience by synthesizing bioindicators, such as forest cover, density, and species richness with geo-indicators, such as fire history, slope, and parent material. The significance of each factor was assessed using sensitivity analysis in order to produce a map of areas at risk- “risk hotspots” – of losing resilience, allowing managers to prioritize resources in restoration efforts.

Kulakowski, D, Veblen, TT, Bigler, Christof, Kulakowski, Dominik, & Veblen, Thomas T. (2005). MULTIPLE DISTURBANCE INTERACTIONS AND DROUGHT INFLUENCE FIRE SEVERITY IN ROCKY MOUNTAIN SUBALPINE FORESTS. Ecology., 86(11), 3018–3029. https://doi.org/10.1890/05-0011

GIS technologies are especially helpful for spatially predicting indicators such as fire severity, which can be the result of complex interactions. This study examines the possible combined effects of interactions between the disturbances of fire, insect outbreaks, and storm blowdown upon fire severity. Pairwise overlay analyses were performed in order to assess these associations. The regression models created, unlike bivariate overlay analysis, allowed for the simultaneous predictions, hypothesis tests and assessment of effects. Results showed that local forest cover type was a significant factor affecting fire spread and severity in the Rocky Mountains, with Spruce-fir stands having the highest probability of burning at high severity. Maps created in GIS show weather variability only significantly affecting fire when fuel build up is sufficient. Pre-fire disturbance and topography were also found to influence burn severity and explain variability.

Michener, W. K. (1997). Quantitatively Evaluating Restoration Experiments: Research Design, Statistical Analysis, and Data Management Considerations. Restoration Ecology : the Journal of the Society for Ecological Restoration., 5(4), 324–337. https://doi.org/10.1046/j.1526-100X.1997.00546.x

Ecological restoration projects (i.e. post-wildfire disturbance) can be very difficult to design and analyze quantitatively due to several factors: experimental units are often heterogeneous, multiple non-uniform treatments may be applied iteratively, replication is difficult or impossible, the effects of extrinsic and intrinsic disturbances may be poorly understood, and the goal of focus is typically the variability in system responses rather than mean responses. This author provides thorough explanations of each of these challenges, along with a variety of ways in which they might be addressed, including via the application of GIS technologies. GIS is described as a powerful tool in relation to this discipline because of its ability to quickly synthesis data using multiple layers, rename and reclassify attributes, analyze spatial coincidence and proximity, and provide quantitative and statistical measurements which can be used to identify potential restoration sites and to visualize and interpret results.

Schroeder, P, Kern, JS, Brown, Sandra L, Schroeder, Paul, & Kern, Jeffrey S. (1999). Spatial distribution of biomass in forests of the eastern USA. Forest Ecology and Management, 123(1), 81–90. https://doi.org/10.1016/S0378-1127(99)00017-1

Biomass is defined by the net difference between photosynthetic production and consumption (respiration, mortality, harvest, herbivory). This measurement is an important indicator of the carbon stored in forests, which can be released as atmospheric carbon into the air during a disturbance (i.e. wildfire) or function as atmospheric carbon sinks during periods of regeneration post-disturbance. While accurate measurements of biomass provide valuable information for decision makers and land managers, large scale estimations can be challenging (remote sensing techniques have met with little success). The authors of this widely-cited study decided to use preciously established methods to convert US forest inventory volume data into above and belowground biomass, downloaded from the USFS Forest Inventory and Analysis (FIA) database, which they then mapped in a geographic information system (GIS) by county. These maps provide a vivid visual representation of forest biomass density patterns over space which can be useful in predicting changes to the global carbon cycle and evaluating potential for increased biomass-carbon storage.

Smith, JE, Weinstein, DA, Laurence, JA, Woodbury, Peter B, Smith, James E, Weinstein, David A, & Laurence, John A. (1998). Assessing potential climate change effects on loblolly pine growth: A probabilistic regional modeling approach. Forest Ecology and Management, 107(1-3), 99–116. https://doi.org/10.1016/S0378-1127(97)00323-X

In this study, a geographic information system was used to integrate regional data including forest distribution, growth rate, and stand characteristics, provided by the USDA Forest Service, with current and predicted climate data in order to produce four different models predicting the potential effects of climate change upon the loblolly pine across the southern U.S. Results indicated a high likelihood of a 19 to 95 percent decrease in growth rates, varying substantially per region and primarily influenced by a relative change in carbon assimilation and CO2 concentrations. In this case, GIS seems particularly useful for synthesizing existing information from regional surveys and account for uncertainties to produce ecological risk assessments at large scales in a way that is useful to policy and decision makers (vs. lengthy reports that are difficult to parse through).

Williams, D, Barry, D, Kasischke, Eric S, Williams, David, & Barry, Donald. (2002). Analysis of the patterns of large fires in the boreal forest region of Alaska. International Journal of Wildland Fire, 11(2), 131–144. https://doi.org/10.1071/WF02023

This study represents the first attempt to spatially correlate the distribution of fire activity in Alaska with climate, topographic, and vegetation cover features using GIS to provide a realistic assignment of fire cycle (frequency) for 11 distinct Alaskan ecoregions (where 96% of all fire activity occurs). GIS technologies can make use of Alaska’s state-wide initiative to digitize maps of fire perimeters from fire events from 1950 to 1999. Perimeter maps are created using a combination of ground and aerial surveys, and aerial photography or satellite imagery. Geospatial analysis showed fire frequency to be influenced by the complex interaction of elevation, aspect, lightening strike frequency, precipitation, forest cover, and growing season temperature.

Please send comments and questions to Melissa Hannah at hannahm@oregonstate.edu or click "Comments Are Welcome" at the top of the page.