<?xml version="1.0" encoding="UTF-8"?><metadata xml:lang="en">
<dataIdInfo>
<envirDesc>Esri ArcGIS 10.6.0.8321</envirDesc>
<dataLang>
<languageCode value="eng"/>
<countryCode value="USA"/>
</dataLang>
<idCitation>
<resTitle>4 ft.Flame-length Exceedance Probability (FLP_GT4)</resTitle>
<presForm>
<PresFormCd value="011"/>
</presForm>
</idCitation>
<spatRpType>
<SpatRepTypCd value="002"/>
</spatRpType>
<dataExt>
<geoEle>
<GeoBndBox esriExtentType="search">
<exTypeCode>1</exTypeCode>
<westBL>-124.868342</westBL>
<eastBL>-116.404817</eastBL>
<northBL>46.350985</northBL>
<southBL>41.910940</southBL>
</GeoBndBox>
</geoEle>
</dataExt>
<idPurp>The purpose of the USFS Pacific Northwest Region Wildfire Risk Assessment (PNRA) is to provide foundational information about wildfire hazard and risk to highly valued resources and assets across the Region. Such information supports regional fuel management planning decisions, as well as revisions to land and resource management plans. A wildfire risk assessment is a quantitative analysis of assets and resources and how they would be potentially impacted by wildfire. The PNRA analysis considers several different components, each resolved spatially across the region, including:
• likelihood of a fire burning, • the intensity of a fire if one should occur,
• the exposure of assets and resources based on their locations, and • the susceptibility of those assets and resources to wildfire FSim – Large-wildfire simulation system
FSim is a comprehensive fire occurrence, growth, behavior, and suppression simulation system that uses locally relevant fuel, weather, topography, and historical fire occurrence information to generate spatially resolved estimates of the contemporary likelihood and intensity of wildfire events (Finney and others 2011). FSim generates stochastic simulation data based on many thousands of iterations, then integrates those iterations into a probabilistic result. An FSim iteration spans one entire year. Due to the highly varied nature of weather and fire occurrence across the Pacific Northwest Region, we divided the landscape into twenty-three Fire Occurrence Areas (FOA). FOA boundaries were generated using elevation-based ecozones, aggregated where appropriate for fire modeling purposes. For consistency with other FSim projects, we numbered these FOAs 401 through 423. Each FOA was simulated independently with a minimum of 10,000 iterations, and then the 23 runs were compiled into a single data product. For each FOA, we parameterized and calibrated FSim based on the location of historical fire ignitions within the FOA, which is consistent with how the historical record is compiled. We then used FSim to start fires only within each FOA, but allowed those fires to spread outside of the FOA. Landscape zones
The Analysis Area (AA) is the area for which valid burn probability (BP) results are produced. The PNRA analysis includes 17 National Forests (NF): Colville, Deschutes, Fremont-Winema, Gifford Pinchot, Malheur, Mt. Baker-Snoqualmie, Mt. Hood, Ochoco, Okanogan-Wenatchee, Olympic, Rogue River-Siskiyou, Siuslaw, Umatilla, Umpqua, Wallowa-Whitman, and Willamette National Forests (NF), as well as the Columbia River Gorge National Scenic Area.
To ensure valid BP results in the AA and prevent edge effects, it is necessary to allow FSim to start fires outside of the AA and burn into it. This larger area where simulated fires are started is called the Fire Occurrence Area (FOA). We established the FOA extent as a 30-km buffer on the Analysis Area. The 30-km buffer provides a sufficient area to make sure that all fires that could reach the AA are simulated. The buffered Fire Occurrence Area covers approximately 116 million acres. Additionally, the available fuelscape extent was determined by adding an additional 30-km buffer to the FOA boundary. This buffer allows fire starting within the FOA to grow unhindered by the edge of the fuelscape, which would otherwise truncate fire growth and affect the simulated fire-size distribution and potentially introduce errors into the calibration process. A map of the Analysis Area, Fire Occurrence Area boundaries and fuelscape extent as well as additional relevant information can be found in the PNRA project report. Fuelscape
The fuelscape consists of geospatial data layers representing surface fuel model, canopy base height, canopy bulk density, and topography characteristics (slope, aspect, elevation). The fuelscape was developed from LANDFIRE 2014 (LF_1.4.0) 30-m raster data and was updated based on USFS resource staff input. Additionally, the fuelscape was updated using RAVG, MTBS, and GeoMac datasets to account for wildfire disturbances that occurred between 2015 and 2017. Weather Data
FSim requires information regarding historic weather. We identified representative weather stations across the landscape that have consistent hourly wind and daily fire weather observations. Weather data from the selected Remote Automated Weather Stations (RAWS) was used to produce monthly distributions of wind speed and direction. For historical ERCs, we used ERC values sampled from Dr. Matt Jolly’s gridded historical ERC rasters. ERCs were sampled at an advantageous location within each FOA. Those locations are found on relative flat ground with little or no canopy cover, in the general area where large-fires have historically occurred. For simulated ERCs in FSim, we used a new feature of FSim that allows the user to supply a stream of ERC values for each FOA. Isaac Grenfell, statistician at the Missoula Fire Sciences Lab, has generated 1,000 years of daily ERC values (365,000 ERC values) on the same 4-km grid as Jolly’s historical ERCs. The simulated ERC values Grenfell produces are “coordinated” in that a given year and day for one FOA corresponds to the same year and day in all other FOAs—their values only differ due to their location on the landscape. This coordination permits analysis of fire-year information across all FOAs. The location of the selected RAWS stations and ERC sample locations as well as more detailed information on how the required weather data products were generated are provided in the PNRA project report. Model Calibration FSim simulations for each FOA were calibrated to historical measures of large fire occurrence including: mean historical large-fire size, mean annual burn probability, mean annual number of large fires per million acres, and mean annual area burned per million acres. From these measures, two calculations are particularly useful for comparing against and adjusting FSim results: 1) mean large fire size, and 2) number of large fires per million acres.
To calibrate each FOA, we started with baseline inputs, and a starting ADJ file informed by experience on previous projects. All runs were completed at 120-m resolution. Each FOA was calibrated separately to well within the 70% confidence interval and final simulations were run with a minimum of 10,000 iterations. The twenty-three FOAs were then integrated using a natural weighting method into an overall result for the analysis area. With this method, well within the boundary of a FOA the results are influenced only by that FOA. Near the border with another FOA the results will be influenced by that adjacent FOA. The weighting of each FOA is in proportion to its contribution to the overall BP at each pixel.
Downscaling FSim results for effects analysis
FSim’s stochastic simulation approach can be computationally intensive and therefore, time constraining on large landscapes. A challenge, therefore, is to determine a resolution sufficiently fine to retain detail in fuel and terrain features, yet produce calibrated results in a reasonable timeframe. Moreover, HVRA are often mapped at the same resolution as the final BP and FLPs produced by FSim. To enable greater resolution on HVRA mapping, we chose to downscale the FSim results to 30 m, consistent with HVRA mapping at 30 m. We downscaled FSim results using a multi-step process. First, we resampled the original, 120-m FSIM result grids to 30 m. Next, we used the Focal Statistics tool in ESRI’s ArcGIS to calculate mean value, of burnable pixels only, within a 7-pixel by 7-pixel moving window. Finally, we used the smoothed values to “backfill” burnable pixels at 30 m that were coincident with non-burnable fuel at 120 m. The final smoothed grids resulted in original FSim values for pixels that were burnable at both 120 m and 30 m, non-zero burn probability values in burnable pixels that were non-burnable at 120m, and a BP of zero in non-burnable, 30 m pixels.
</idPurp>
<idAbs>&lt;DIV STYLE="text-align:Left;"&gt;&lt;DIV&gt;&lt;DIV&gt;&lt;P STYLE="margin:0 0 11 0;"&gt;&lt;SPAN&gt;&lt;SPAN&gt;This dataset represents the conditional probability of exceeding a nominal flame-length value (also known as flame-length exceedance probability, or FLEP). There are five FLEP rasters. FLEP_GT2 is the conditional probability of exceeding a flame length of 2 feet; it is calculated as the sum of iFLP_FIL2 through iFLP_FIL6. FLEP_GT4 is the conditional probability of exceeding a flame length of 4 feet; it is calculated as the sum of iFLP_FIL3 through iFLP_FIL6. FLEP_GT6 is the conditional probability of exceeding a flame length of 6 feet; it is calculated as the sum of iFLP_FIL4 through iFLP_FIL6. FLEP_GT8 is the conditional probability of exceeding a flame length of 8 feet; it is calculated as the sum of iFLP_FIL5 and iFLP_FIL6. There is no raster for FLEP_GT0 because, by definition, for all burnable pixels there is a 100 percent probability that flame length will exceed 0, given that a fire occurs. &lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/P&gt;&lt;P STYLE="margin:0 0 11 0;"&gt;&lt;SPAN&gt;The iFLP_FILx rasters are the integrated (project wide) conditional probabilities of observing flame length in each of six classes: iFLP_FIL1 represents flame lengths from 0 - 2 ft, iFLP_FIL2 represents flame lengths from 2 - 4 ft, iFLP_FIL3 represents flame lengths from 4 - 6 ft, iFLP_FIL4 represents flame lengths from 6 - 8 ft, iFLP_FIL5 represents flame lengths from 8 - 12 ft, and iFLP_FIL6 represents flame lengths &amp;gt;12 ft.&lt;/SPAN&gt;&lt;/P&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;/DIV&gt;</idAbs>
<idCredit>This dataset was developed for the USFS Pacific Northwest Region by Pyrologix LLC (www.pyrologix.com).</idCredit>
<resConst>
<Consts>
<useLimit>&lt;DIV STYLE="text-align:Left;"&gt;&lt;DIV&gt;&lt;DIV&gt;&lt;P STYLE="margin:0 0 11 0;"&gt;&lt;SPAN&gt;&lt;SPAN&gt;The user must be aware of data conditions and must ultimately bear responsibility for the appropriate use of the information with respect to possible errors, possible omissions, map scale, data collection methodology, data currency, and other conditions specific to certain data.&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN /&gt;&lt;/P&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;/DIV&gt;</useLimit>
</Consts>
</resConst>
</dataIdInfo>
<mdLang>
<languageCode value="eng"/>
<countryCode value="USA"/>
</mdLang>
<mdChar>
<CharSetCd value="004"/>
</mdChar>
<distInfo>
<distFormat>
<formatName>File Geodatabase Raster Dataset</formatName>
</distFormat>
</distInfo>
<mdHrLv>
<ScopeCd value="005"/>
</mdHrLv>
<mdHrLvName>dataset</mdHrLvName>
<refSysInfo>
<RefSystem>
<refSysID>
<identCode code="2992"/>
<idCodeSpace>EPSG</idCodeSpace>
<idVersion>8.9.3(9.0.0)</idVersion>
</refSysID>
</RefSystem>
</refSysInfo>
<spdoinfo>
<rastinfo>
<rasttype>Pixel</rasttype>
<rowcount>16168</rowcount>
<colcount>21744</colcount>
<vrtcount>1</vrtcount>
<rastxsz Sync="TRUE">98.425197</rastxsz>
<rastysz Sync="TRUE">98.425197</rastysz>
<rastbpp Sync="TRUE">32</rastbpp>
<rastorig Sync="TRUE">Upper Left</rastorig>
<rastcmap Sync="TRUE">FALSE</rastcmap>
<rastcomp Sync="TRUE">LZ77</rastcomp>
<rastband Sync="TRUE">1</rastband>
<rastdtyp Sync="TRUE">pixel codes</rastdtyp>
<rastifor Sync="TRUE">FGDBR</rastifor>
<rastplyr Sync="TRUE">TRUE</rastplyr>
</rastinfo>
</spdoinfo>
<spref>
<horizsys>
<planar>
<planci>
<plance>row and column</plance>
<coordrep>
<absres>98.425197</absres>
<ordres>98.425197</ordres>
</coordrep>
</planci>
</planar>
</horizsys>
</spref>
<mdDateSt>20180410</mdDateSt>
<Esri>
<DataProperties>
<itemProps>
<itemName Sync="TRUE">Probability_of_exceeding_4ft_flames</itemName>
<imsContentType Sync="TRUE">002</imsContentType>
<nativeExtBox>
<westBL Sync="TRUE">914844.429643</westBL>
<eastBL Sync="TRUE">1472127.894210</eastBL>
<southBL Sync="TRUE">678506.758930</southBL>
<northBL Sync="TRUE">1029884.711686</northBL>
<exTypeCode Sync="TRUE">1</exTypeCode>
</nativeExtBox>
</itemProps>
<coordRef>
<type Sync="TRUE">Projected</type>
<geogcsn Sync="TRUE">GCS_North_American_1983</geogcsn>
<csUnits Sync="TRUE">Linear Unit: Foot (0.304800)</csUnits>
<projcsn Sync="TRUE">NAD_1983_Oregon_Statewide_Lambert_Feet_Intl</projcsn>
<peXml Sync="TRUE">&lt;ProjectedCoordinateSystem xsi:type='typens:ProjectedCoordinateSystem' xmlns:xsi='http://www.w3.org/2001/XMLSchema-instance' xmlns:xs='http://www.w3.org/2001/XMLSchema' xmlns:typens='http://www.esri.com/schemas/ArcGIS/2.7.0'&gt;&lt;WKT&gt;PROJCS[&amp;quot;NAD_1983_Oregon_Statewide_Lambert_Feet_Intl&amp;quot;,GEOGCS[&amp;quot;GCS_North_American_1983&amp;quot;,DATUM[&amp;quot;D_North_American_1983&amp;quot;,SPHEROID[&amp;quot;GRS_1980&amp;quot;,6378137.0,298.257222101]],PRIMEM[&amp;quot;Greenwich&amp;quot;,0.0],UNIT[&amp;quot;Degree&amp;quot;,0.0174532925199433]],PROJECTION[&amp;quot;Lambert_Conformal_Conic&amp;quot;],PARAMETER[&amp;quot;False_Easting&amp;quot;,1312335.958005249],PARAMETER[&amp;quot;False_Northing&amp;quot;,0.0],PARAMETER[&amp;quot;Central_Meridian&amp;quot;,-120.5],PARAMETER[&amp;quot;Standard_Parallel_1&amp;quot;,43.0],PARAMETER[&amp;quot;Standard_Parallel_2&amp;quot;,45.5],PARAMETER[&amp;quot;Latitude_Of_Origin&amp;quot;,41.75],UNIT[&amp;quot;Foot&amp;quot;,0.3048],AUTHORITY[&amp;quot;EPSG&amp;quot;,2992]]&lt;/WKT&gt;&lt;XOrigin&gt;-118489100&lt;/XOrigin&gt;&lt;YOrigin&gt;-97381100&lt;/YOrigin&gt;&lt;XYScale&gt;37592196.316242374&lt;/XYScale&gt;&lt;ZOrigin&gt;-100000&lt;/ZOrigin&gt;&lt;ZScale&gt;10000&lt;/ZScale&gt;&lt;MOrigin&gt;-100000&lt;/MOrigin&gt;&lt;MScale&gt;10000&lt;/MScale&gt;&lt;XYTolerance&gt;0.0032808398950131233&lt;/XYTolerance&gt;&lt;ZTolerance&gt;0.001&lt;/ZTolerance&gt;&lt;MTolerance&gt;0.001&lt;/MTolerance&gt;&lt;HighPrecision&gt;true&lt;/HighPrecision&gt;&lt;WKID&gt;2992&lt;/WKID&gt;&lt;LatestWKID&gt;2992&lt;/LatestWKID&gt;&lt;/ProjectedCoordinateSystem&gt;</peXml>
</coordRef>
</DataProperties>
<SyncDate>20210618</SyncDate>
<SyncTime>08380000</SyncTime>
<ModDate>20210618</ModDate>
<ModTime>08380000</ModTime>
</Esri>
</metadata>
