Analyzing Wildfire Risk and Endangered Species in Trinity County Using ArcGIS Pro

Summary

The goal of our model is to identify areas of high fire risk in Trinity County, California and where they intersect with endangered species' critical habitats. Using 9 different variables, a multicriteria raster model was created to determine fire ignition risk. This will allow fire risk in critical habitat areas to be mitigated in the future to protect endangered species. This project was created in collaboration with partner Serena Morgan, a CU boulder student.

Research Questions

How do areas of high wildfire risk correspond with endangered species habitats?
Where are the areas with the highest wildfire risk?
Which endangered species’ habitats are most at risk?

Background

Fire risk is an ongoing problem in the United States and has increased in recent years, especially with climate change creating more prevalent droughts and increasing temperatures (National Oceanic and Atmospheric Administration). In this model, we define fire risk as the likelihood of fire ignition in a given location.

Study Area

Our study area is Trinity County, a rural county located in northern California. It has a population of 16,000 and has no incorporated communities. It is a mountainous county, with an average elevation of 868 meters and a maximum elevation of 2754 meters (“Trinity County”; “Trinity County Topographic Map, Elevation, Terrain” ). The predominant land cover is evergreen forest, grassland, and shrub (Multi-Resolution Land Characteristics Consortium). It has a Mediterranean climate, with warm dry summers and cool wet winters. The average annual precipitation is 37.5 inches, and the average annual temperature is 55º Fahrenheit (“Trinity County”). Since 1998, 40% of Trinity County has been affected by wildfires (Wesley). Additionally, 3 endangered species critical habitats exist in Trinity County: the Northern Spotted Owl, the Marbled Murrelet, and the Lassics Lupine. The Northern Spotted Owl habitat is by far the largest in the county, with the Marbled Murrelet and Lassics Lupine covering very small portions of the county (California Energy Commission).

Data Collection

Data were acquired from free spatial data sources available online.

Preprocessing: Variable Creation

This data was then processed for use in the model. First, we converted the precipitation and temperature data from a NetCDF file format to a raster file format. We also used “mosaic to new raster” to generate a seamless DEM for Trinity County. From this DEM, aspect and slope rasters were created. Using the Euclidean distance tool, a raster was created that showed distance from roads. We used Sentinel 2 bands 8a and 11 to create an NDMI raster by inputting the equation (B8A - B11)/(B8A + B11) into the raster calculator (“Normalized Difference Moisture Index: Equation and Interpretation”). Using select by attribute, fire history polygons were filtered to only include fires that occurred since the year 2000. Then, these polygons were converted to raster. Lightning point data was used to create a kernel density raster to show where lightning strikes occur most. Each dataset for the multicriteria raster model was resampled to 30m resolution, reprojected to California State Plane Zone 1: NAD 1983, and clipped to the Trinity County boundary.

Model

The model for the multi criteria raster model started with reclassifying each of the rasters from a 1 to 6 scale, with 6 being highest fire risk and 1 being lowest. The variables were divided into 2 categories: Topographical variables, and temporal variables. Topograpahic variables include Euclidean distance to roads, slope, aspect, land cover, and fire history. Temporal variables include average temperature, precipitation, and soil moisture for September 2024. Lightning data was from August 2020, as no data could be found for September 2024.

Reclassification

After preprocessing, the first step was to reclassify each raster to a common scale using research to determine which factors elevate the risk for fire ignition. 6 classes were used, as this was the number of classes that could be made for the aspect raster. A value of 6 represents the highest fire risk, while a value of 1 represents the lowest fire risk. We did extensive research to justify our slope classification and generally found that while steeper slopes are more prone to fire spread, flatter slopes are more prone to fire risk (Taylor et al.; Harris et al.; Salavati et al.; Mohammed et al.; Baltaci et al.; Maingi and Henry). Abdo et al. (2022) and Estes et al. (2017) used this aspect classification for their fire studies in Syria and Northern California, respectively, with east and south facing slopes being most fire prone. We justified our land cover classification using information from Butler et al. (2007) and a wildfire spread report from the Alberta Government, with grassland and pasture being at highest risk. Fire history was justified with a study by (Taylor et al.), with newer burn scars being less fire prone. For distance from roads, Maingi and Henry (2007) included roads in their fire occurrence and distribution model and found that distance to roads was a prominent factor in wildfire occurrence, with fires more likely to start closer to roads. For NDMI (soil moisture) For lightning kernel density, a higher desnity of strikes increases fire ignition risk. For precipitation, less average precipitation increases fire ignition. For temperature, higher temperature will increase the fire ignition risk.

Raster Variables After Reclassification

Weighted Sum

After reclassifying each raster, the weighted sum tool was used to create a fire risk raster for Trinity County. The reclassified rasters were split into two categories: topological variables (road Euclidean distance, slope, aspect, land cover, and fire history) and weather-related factors (temperature, precipitation, NDMI, and lightning strike density). Weather related variables were obtained for September 2024, besides lighting density (August 2020). For topological variables, road Euclidean distance was given a weight of 2, slope 1, aspect 1, land cover 1.5, and fire history 2 (Padilla and Vega-García; Maingi and Henry). For weather-related variables, temperature was given a weight of 1, precipitation 2, soil moisture 2, and lightning density 2 (Padilla and Vega-García). These two rasters were then combined to create a weighted sum raster that included both weather and topological variables.

Distance from roads, land cover, and fire history were given greater importance in the model.

Precipitation, NDMI, and lightning density was given increased importance in the model.

Topographic variables and weather related raster models added.

The Final Fire Risk Model

Second Part of Model

The second part of the model after the the fire risk model involves using zonal statistics to determine which critical habitat areas have the highest fire risk. It starts with using the locate regions tool to find the three areas with highest fire risk. Then, Critical habitat areas were overlaid on these regions to find the highest risk critical habitats. Other zonal statistics were run to find the average value of all of Trinity county, and the locate regions output.

Locate Regions

The locate regions tool was used on the fire risk raster model in order to find three areas with heightened fire risk.

Critical Habitats

Here are critical habitats for Trinity County. These are "specific geographic area(s) that contains features essential for the conservation of a threatened or endangered species and that may require special management and protection" (California Energy Commission).

Extract by Mask

Using extract by mask, the cititical habitat regions within the highest fire risk zones could be delineated.

Results

Using zonal statistics, information about the fire risk model could be obtained. Information about the whole of Trinity County, the three highest fire risk regions, and the critical habitat areas with the highest fire risk were acquired from zonal statistics.

The Northern Spotted Owl

The endangered species most at risk is the Northern Spotted Owl. It is the only species whose habitat is within the highest fire risk areas. This species is classified as “threatened” and has an annual population decline of about 3% (American Bird Conservancy). Primary threats include competition with the invasive Barred Owl and habitat loss due to logging and wildfire (American Bird Conservancy; “Northern Spotted Owls in California; U.S. Geological Survey, “Northern Spotted Owl”). It is an old-growth forest obligate and is unable to live elsewhere (American Bird Conservancy; “Northern Spotted Owls in California; U.S. Geological Survey). Other endangered species in Trinity County are the Marbled Murrelet and the Lassics Lupine, but they are not located in the highest fire risk regions.

Conclusion

Using ArcGIS pro, a fire risk model was created for Trinity County. Using nine variables in raster format, the weighted sum tool was used to give each variable more or less importance in the model based on fire risk research. Three regions of high fire risk were identified using the locate regions tool. Critical habitat areas were overlaid on the risk model and the highest risk habitats were identified.

Overall, Trinity County is at a very elevated risk for fire. The presence of the Northern Spotted Owl habitat, especially in the highest fire risk zones we identified, must be addressed by implementing fire mitigation measures, which can include forest thinning and controlled burns. This will help mitigate the chances of a severe burn occurring in the Northern Spotted Owl habitat to keep their ecosystems intact.

Works Cited

Abdo, Hazem Ghassan, et al. “GIS-Based Frequency Ratio and Analytic Hierarchy Process for Forest Fire Susceptibility Mapping in the Western Region of Syria.” Sustainability, vol. 14, no. 8, 13 Apr. 2022, p. 4668, https://doi.org/10.3390/su14084668.

Alberta Government, Agriculture and Irrigation. How different tree species impact the spread of wildfire. Department Documents, 3 August 2012. www1.agric.gov.ab.ca/$department/deptdocs.nsf/all/formain15744/$FILE/tree-species-impact-wildfire-aug03-2012.pdf.

American Bird Conservancy. “Northern Spotted Owl.” American Bird Conservancy, 2010, abcbirds.org/bird/northern-spotted-owl/. Accessed Nov. 2024.

Baltaci, Uğur, and Feriha Yildirim. “Effect of Slope on the Analysis of Forest Fire Risk.” Hacettepe Journal of Biology and Chemistry, vol. 48, no. 4, 3 Sept. 2020, pp. 373–379, https://doi.org/10.15671/hjbc.753080.

Butcher, Kristin. Wildfire Exposure to Critical Habitat of Endangered and Threatened Species in California. 18 May 2019, repository.usfca.edu/thes/1322.

Butler, B.W.; Anderson, W.R.; Catchpole, E.A. 2007. Influence of slope on fire spread rate. In: Butler, Bret W.; Cook, Wayne, comps. The fire environment--innovations, management, and policy; conference proceedings. 26-30 March 2007; Destin, FL. Proceedings RMRS-P-46CD. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. CD-ROM. p. 75-82

California Department of Forestry and Fire Protection. California Fire Perimeters (all). 29 August 2024. Feature Layer. https://catalog.data.gov/dataset/california-fire-perimeters-all-5b4ee. Accessed Nov. 2024.

California Energy Commission. California Counties. 14 February 2023. Shapefile. https://catalog.data.gov/dataset/california-counties-254c9. Accessed Nov. 2024.

California Energy Commission. Critical Habitat. 6 March 2023. Shapefile. https://catalog.data.gov/dataset/fws-critical-habitat-for-threatened-and-endangered-species-dataset. Accessed Oct. 2024.

“California State Plane Coordinate System.” California Department of Conservation, www.conservation.ca.gov/cgs/rgm/state-plane-coordinate-system. Accessed Nov. 2024.

Estes, Becky L., et al. “Factors Influencing Fire Severity under Moderate Burning Conditions in the Klamath Mountains, Northern California, USA.” Ecosphere, vol. 8, no. 5, May 2017, p. e01794, https://doi.org/10.1002/ecs2.1794.

FEMA Region 9. “California Lightning Strikes Recorded from August 16-18, 2020. 22 Aug. 2020. Feature Layer. https://www.arcgis.com/home/item.html?id=7adfdfc4325447ef9a64b4f925b3ee93. Accessed Dec. 2024.

Harris, Lucas, and Alan H. Taylor. “Previous Burns and Topography Limit and Reinforce Fire Severity in a Large Wildfire.” Ecosphere, vol. 8, no. 11, Nov. 2017, p. e02019, https://doi.org/10.1002/ecs2.2019.

Maingi, John K., and Mary C. Henry. “Factors Influencing Wildfire Occurrence and Distribution in Eastern Kentucky, USA.” International Journal of Wildland Fire, vol. 16, no. 1, 2007, p. 23, https://doi.org/10.1071/wf06007.

Mohammed, Osama Ashraf, et al. “A Comparative Study of Forest Fire Mapping Using GIS-Based Data Mining Approaches in Western Iran.” Sustainability, vol. 14, no. 20, 21 Oct. 2022, p. 13625, https://doi.org/10.3390/su142013625.

Multi-Resolution Land Characteristics Consortium. USA NLCD Land Cover. 2021. GeoTIFF. https://www.arcgis.com/home/item.html?id=3ccf118ed80748909eb85c6d262b426f. Accessed Nov. 2024.

National Oceanic and Atmospheric Administration. “Wildfire Climate Connection.” National Oceanic and Atmospheric Administration, 8 Aug. 2022, www.noaa.gov/noaa-wildfire/wildfire-climate-connection. Accessed Nov. 2024.

National Oceanic and Atmospheric Administration and National Centers for Environmental Information. Continuous US Average Monthly Precipitation. 11 Dec. 2024. NetCDF. https://www.drought.gov/data-maps-tools/gridded-climate-datasets-noaas-nclimgrid-monthly. Accessed Nov. 2024.

National Oceanic and Atmospheric Administration and National Centers for Environmental Information. Continuous US Average Monthly Temperature. 11 Dec. 2024. NetCDF. https://www.drought.gov/data-maps-tools/gridded-climate-datasets-noaas-nclimgrid-monthly. Accessed Nov. 2024.

“Normalized Difference Moisture Index: Equation and Interpretation.” EOS Data Analytics, 7 Feb. 2022, eos.com/make-an-analysis/ndmi/. Accessed Nov. 2024.

“Northern Spotted Owls in California.” California Department of Fish and Wildlife, wildlife.ca.gov/Conservation/Birds/Northern-Spotted-Owl. Accessed Nov. 2024.

Padilla, M., and C. Vega-García. “On the Comparative Importance of Fire Danger Rating Indices and Their Integration with Spatial and Temporal Variables for Predicting Daily Human-Caused Fire Occurrences in Spain.” International Journal of Wildland Fire, vol. 20, no. 1, 2011, p. 46, https://doi.org/10.1071/wf09139.

Pramatana, F, et al. “The Habitat Susceptibility of Bali Starling (Leucopsar RothschildiStresemann> 1912) Based on Forest Fire Vulnerability Mappin in West Bali National Park.” IOP Conference Series: Earth and Environmental Science, vol. 91, no. 012003, Oct. 2017, p. 012003, https://doi.org/10.1088/1755-1315/91/1/012003.

Salavati, Ghafar, et al. “Wildfire Risk Forecasting Using Weights of Evidence and Statistical Index Models.” Sustainability, vol. 14, no. 7, 25 Mar. 2022, p. 3881, https://doi.org/10.3390/su14073881.

Sentinel-2. Band 8a (Narrow Near Infrared, 855-875 nm) and Band 11 (Short Wave Infrared, 1565-1655 nm). TIFF. 13 Sep. 2024. https://apps.sentinel-hub.com/eo-browser/. Accessed Nov. 2024.

Taylor, Alan H., et al. “Drivers of Fire Severity Shift as Landscapes Transition to an Active Fire Regime, Klamath Mountains, USA.” Ecosphere, vol. 12, no. 9, Sept. 2021, https://doi.org/10.1002/ecs2.3734.

“Trinity County.” Wikipedia, Wikimedia Foundation, 30 Dec. 2019, en.wikipedia.org/wiki/Trinity_County. Accessed Nov. 2024

“Trinity County Topographic Map, Elevation, Terrain.” Topographic Maps, 2024, en-ca.topographic-map.com/map-s7ndf3/Trinity-County/?center=41.49569%2C-121.06483&zoom=7. Accessed Nov. 2024.

“Trinity Forest Health and Fire-Resilient Rural Communities.” Arcgis.com, 2024, experience.arcgis.com/experience/0fe032e92fad464fbcdc7faf12cd7928/page/-Trinity-Forest-Health-and-Fire-Resilient-Rural-Communities/. Accessed Nov. 2024.

U.S. Geological Survey. 1 Arc Second DEM (n40w123; n40w124; n41w123; n41w124; n42w123; n42w124). 13 March 2024. GeoTIFF. https://apps.nationalmap.gov/downloader/. Accessed Oct. 2024.

U.S. Geological Survey. “Northern Spotted Owl Still Fights for Survival.” U.S. Geological Survey, 6 Oct. 2021, www.usgs.gov/news/featured-story/northern-spotted-owl-still-fights-survival. Accessed Nov. 2024.

Wesley, Denise. Major Roads of Trinity County, CA. 7 May 2024. Feature Layer. https://www.arcgis.com/home/item.html?id=f414a1fd769f4ce6a2d9172a3f08ab58. Accessed Nov. 2024.

Wesley, Denise. “Trinity County Wildfire History Timeline.” ArcGIS StoryMaps, Esri, 25 June 2024, storymaps.arcgis.com/stories/34908fe7b97247afa5269559c64b04a9. Accessed Nov. 2024.

For more information about this project, view the full project report here.

Page updated

Google Sites

Report abuse