3DSYCANGRASS_2026-03-19_1501 SusanPrichardUniveristy of WashingtonResearch Scientist

Blodel Hall 361University of Washington Seattle WA 98195 United States

sprich@uw.eduhttps://depts.washington.edu/flame/ https://orcid.org/0000-0002-6001-1487 EricRowellCalifornia Air Resources BoardPrescribed and Agricultural Fire Emissions Specialist

1001 I Street Sacramento California 95814 United States

Eric.Rowell@arb.ca.govhttps://orcid.org/0000-0002-8905-3621 JonathanBatchelorUniveristy of MonanaTeaching Professorjonathan.batchelor@umontana.eduhttps://orcid.org/0000-0001-8700-3076 contributor MichelleBesterSandia National LaboratoryResearchermbester@sandia.gov contributor BenjaminBrightUSFS Rocky Mountain Research StationResearch Scientistbenjamin.c.bright@usda.govhttps://orcid.org/0000-0002-8363-0803contributor JamesCronanUSFS PNW Research StationResearch Forester

400 N 34th St #201 Seattle Washington 98103 United States

jcronan@usda.gov contributor BrianDryeUniveristy of WashingtonSoftware Developer

Bloedel Hall 361University of Washington Seattle Washington 98195 United States

bdrye@uw.edu contributor PaigeEagleUniveristy of WashingtonResearch Coordinator

Bloedel Hall 361University of Washington Seattle Washington 98195 United States

pceagle@uw.eduhttps://depts.washington.edu/flame/https://orcid.org/0009-0003-2894-0037 contributor ChristieHawleyUSFS Southern Research StationResearch Scientistchristie.m.hawley@usda.govhttps://orcid.org/0000-0001-9105-2065contributorJesseJohnsonUniversity of MontanaProfessorjesse.johnson@umontana.eduhttps://orcid.org/0000-0001-7688-2300contributorAndrewHudakUSFS Rocky Mountain Research StationResearch Scientistahudak@fs.fed.ushttps://orcid.org/0000-0001-7480-1458contributorMaureenKennedyUniversity of WashingtonProfessormkenn@uw.eduhttps://orcid.org/0000-0003-4670-3302contributorE. LouiseLoudermilkUSFS Southern Research StationResearch Scientistelloudermilk@fs.fed.ushttps://orcid.org/0000-0001-8191-8670contributor DeborahNemensUniveristy of WashingtonResearch Scientist

Bloedel Hall 361University of Washington Seattle WA 98195 United States

dnemens@uw.eduhttps://depts.washington.edu/flame/https://orcid.org/0000-0002-4388-1923 contributor RogerOttmarUSFS PNW Research StationResearch Foresterroger.ottmar@usda.govhttps://orcid.org/0000-0002-4385-4052contributorRussParsonsUSFS Rocky Mountain Research StationResearch Scientistrparsons@fs.fed.ushttps://orcid.org/0000-0001-5091-8993contributor CarlosSilvaUniversity of FloridaAssistant Professor of Quantitative Forest Science

138 Newins-Ziegler HallUniversity of Florida Gainsville Florida 32611 United States

c.silva@ufl.edu contributor JessicaThoresonUniveristy of WashingtonResearch Scientist

Bloedel Hall 361University of Washington Seattle Washington 98195 United States

jessthor@uw.eduhttps://depts.washington.edu/flame/https://orcid.org/0009-0006-6418-9414 contributor The 3-Dimensional (3D) Fuels study (SERDP Project RC19-1064) integrates field sampling, remote sensing, and 3D modeling to produce 3D fuels datasets for fire, fuels and smoke management mapping applications and wildland fire behavior and smoke modeling. This dataset includes raw, processed, and summarized clip plot field data, forest floor field data, close-range Structure from Motion (SfM) photogrammetry, Terrestrial Laser Scanning (TLS) clip plot datasets, TLS plot datasets, TLS unit datasets, Unmanned Aerial System SfM photogrammetry, and Aerial Laser Scanning (ALS) datasets for the Sycan Grass field site (Sycan Marsh Preserve Grassland Unit, The Nature Conservancy, Oregon). This site-specific dataset requires the metadata, including detailed methods and data dictionaries, that are located in the general dataset "3-Dimensional (3D) Fuels Project: July 2018-February 2024" (see Annotation section for hasMetadata link). The latest version of the general dataset should be used. Similar data were collected at 17 other field sites, and are available in their own site-specific dataset. wildland fuel characteristicsstructure-from-motion photogrammetrypoint cloudsterrestrial laser scanningairborne laser scanningprescribed fireNonefuel heightfuel coverreflectancefuel moistureoccupied volumeforest floorbulk densitylitterduffData Types and Measurement Variables This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Sycan Grass - Sycan Marsh Preserve Grassland Unit, The Nature Conservancy, Oregon. See site-specific dataset. -121.1-121.142.742.7 2019-09-252019-10-12https://datacite-metadata-schema.readthedocs.io/en/4.5/appendices/appendix-1/relationType/#hasMetadatahttps://doi.org/10.60594/W4W016https://schema.org/archivedAthttps://app.globus.org/file-manager?origin_id=095b965e-aa6f-4774-83f8-bf49706d753c&origin_path=/3DSYCANGRASS_2026-03-19_1501/https://schema.org/archivedAthttps://tier2.wfsi-data.org/3DSYCANGRASS_2026-03-19_1501/ SusanPrichard

Bloedel Hall 361University of Washington Seattle WA 98195 United States

sprich@uw.edu https://orcid.org/0000-0002-6001-1487 EXPERIMENTAL DESIGN and DATA COLLECTED. The study site is a 200mx200m area, with a sampling grid at 50m spacing (n=25 grid point locations, see Plot_Set_Up document). Site was sampled in a hierarchical framework, from a 200x200m synoptic to a 10cm voxel level. Data collected at the synoptic scale include TLS (Terrestrial Laser Scanning) synoptic scanning, and high-altitude UAS photogrammetry. Airborne laser scanning (ALS) point cloud data was acquired from other sources and used in analyses. Of the 25 grid point locations, 18 were randomly chosen for higher resolution scanning in 5x5 plots. Data collected at the 5x5 plot scale included TLS scanning (4 scans per plot, one at each corner) and low-altitude UAS photogrammetry (see UAS_Photogrammetry_Sampling document). For 9 of the 5x5m plots, 5 sub-plots were established (center plus cardinal directions within each 5x5m plot) for further sampling. Within these 0.5x0.5x1m subplots (termed "clip plots"), close-range photogrammetry (using a hand-held video camera, see Close_Range_Photogrammetry document), high-resolution photographs, ocular presence-absence (P/A) and destructive sampling of biomass (see Presence_Absence_Survey document) was conducted following Hawley et al. (2018). Destructive sampling of biomass (see Presence_Absence_Biomass_Sampling Survey document) was conducted in 23 of the clip plots before prescribed burns (labeled PRE_D), while the other 22 had destructive sampling only after the prescribed burn was completed (labeled PRE_ND before burn, and POST after burn). See Site Description document in dataset folder for details. Sampling of the forest floor (litter and duff) occurred in the voxel plots (biomass was sampled down to mineral soil in each plot). No additional sampling for litter and duff depth and bulk density was conducted.See methods documents in the general dataset "3-Dimensional (3D) Fuels Project: July 2018-February 2024" (see Annotation section for hasMetadata link. Use the latest version.) for full details: Plot Setup, UAS Photogrammetry, Close-range Photogrammetry sampling, Presence_Absence_Survey.Hawley, C.M., Loudermilk, E.L., Rowell, E.M. and Pokswinski, S., 2018. A novel approach to fuel biomass sampling for 3D fuel characterization. MethodsX, 5, pp.1597-1604. INSTRUMENTS and CALIBRATION. Terrestrial Laser Scanning (TLS): RIEGL -VZ2000 and RIEGL-VZ400i with 750 kHz for the laser sampling rate and vertical sampling density of 0.21 degreess, mounted on a tripod at a minimum height of 1.5 m. The REIGL has an integrated L1 GPS receiver, compass and inclination sensors that aid in geo-referencing point clouds. UAS photogrammetry: UAS-mounted Zenmuse X3 optimal camera and a Micasense Red Edge 3 multispectral sensor. Ground control points were collected as part of the field data collection and were marked with printed coded black and white targets. Coordinates of these were recorded using an RTK/Differential GPS receiver/base combination. Most ground control points were put precisely at the NW corners of the smaller field plots in order to tie the later close-range drone acquisitions to the 200x200m acquisition. High-altitude imagery collected with light parameters include an altitude of 60-100m AGL; with 24 flight lines North-South (NS) and 26 EW to ensure there was ~80% end lap and ~75% side lap for the Micasense. Low-altitude imagery was collected at an altitude of approximately 3m when canopy conditions around the 5x5m plots allowed. Close-range photogrammetry: GoPro Hero7 camera mounted on a Feiyu G6 gimbal which was then threaded onto a 36 inch lightweight rigid pole; a 1-2 minute video was filmed from a constant 1 m elevation and consistent snaking pattern above each voxel plot at 2700K resolution. A second snaking pattern was repeated at a 90° angle to the first, and finally, the entire plot was circled with the camera at an oblique angle pointed inward to the plot. Cylindrical foam flotation devices were placed at random near the edges of each plot; these were used as alignment markers in the imagery to facilitate stitching of individual frames from the video recording. High-resolution photographs were taken (using a Nikon SLR camera) of each voxel plot; included in the frame was a small whiteboard with the site name, plot number, and date. Duff+soil analysis: PerkinElmer 2400 Series II CHNS/O Elemental Analyzer was used to determine carbon mass of mixed duff+soil samples, then percent organic matter (%OM) was calculated from this value (2*%C = %OM) and applied to the dry duff+soil weight for a total pure duff weight.See methods documents in the general dataset "3-Dimensional (3D) Fuels Project: July 2018-February 2024" (see Annotation section for hasMetadata link. Use the latest version.) for full details: UAS Photogrammetry, Close-range Photogrammetry sampling, Duff Processing DATA PROCESSING and QAQC. Raw scans from TLS acquisitions were registered, cleaned of noise points, and processed into laz files using proprietary software (see TLS Processing document). Tiles were created from synoptic scans to cover the entire site. High resolution scans were registered, cleaned of noise points, and processed into point clouds using the same methodology as for synoptic scans. High-resolution point clouds were then clipped down to the 0.5x0.5m voxel plot dimensions for each of the 45 subplots using a shiny app tool developed specifically for this project in R (see TLS_Point_Cloud_Clipping_Shiny_Tool_Instructions document). The clipped laz file was normalized and points were classified as ground/nonground using LAStools software (lasheight and lasground functions). QAQC of lidar point clouds: a random subset of plots were visually inspected and compared with clipped photogrammetry point clouds and high-resolution photos. Additionally, voxel occupancy of point clouds was compared to P/A data (see TLS_Point_Cloud_Clipping_QAQC document). Photogrammetry point clouds were created from a selection of still images extracted from the videos taken in the field using a combination of R, Python and Agisoft scripts developed specifically for this project. Photogrammetry QAQC: visual comparisons of lidar point clouds and high-resolution photos taken of voxel plots in the field were used to assure accuracy. Voxel biomass processing: Biomass samples were shipped to the lab, unpacked and allowed to airdry following field campaigns. Samples from each upper strata (10-100 cm) were not sorted, and were dried and weighed as a unit (e.g. 30-40 cm), while samples from the 0-10 cm strata were sorted into fuel element categories and sub-categories (termed vegcodes) before drying and weighing (see Biomass_Processing document). Litter and duff samples were treated as below. Litter samples were oven-dried without sieving. Duff samples were sieved using standard soils sieves and separated into duff and duff+soil samples, with large soil aggregates and rocks discarded. Pure duff was dried and weighed, and material that fell through the sieve was dried and weighed, labeled as "duff+soil", then sent to a soils lab for analysis of organic matter content. All samples were oven-dried at 70°C for 72 hrs (see Duff_Sample_Processing document). All field and lab data were entered into a database specifically designed to minimize errors, and a technician separate from the one that entered the data reviewed electronic data in a QAQC process (see Biomass_Processing_QAQC document).See methods documents in the general dataset "3-Dimensional (3D) Fuels Project: July 2018-February 2024" (see Annotation section for hasMetadata link. Use the latest version.) for full details: TLS Processing, TLS_Point_Cloud_Clipping_Shiny_Tool_Instructions, TLS Point Cloud Clipping QAQC, Photogrammetry Processing, Duff Processing, Biomass Processing, Presence/Absence Survey QAQC, Biomass Processing QAQC, 3D Fuels Data Dictionary. Roger Ottmar USDA FS PNW Research Station roger.ottmar@usda.gov https://orcid.org/0000-0002-4385-4052 Principal Investigator U. S. Department of Defense (DoD), Strategic Environmental Research and Development Program (SERDP) http://dx.doi.org/10.13039/100013316 RC19-1064