Click the map to explore an area.
About this map
Detailed surfaces show the unchanged native 10 cm model labels, not whole-square acceptance. Purple is uncertain classification; orange is partial photo support; transparent areas have no photo. Colours blend in overview thumbnails. Pixel size is not independent accuracy. No training, threshold change, filling or new classification was performed. The separate 30 m cover-percentage layer retains its original quality gates.
30 m display squares share the 2026 photo origin in EPSG:32613. This is an aggregation scale, not proof of matching the 2021 CWPP definition. Absolute photo placement is not survey-qualified.
2026 woody cover: percentage of readable area covered by tree and shrub crowns. Original frozen exploratory classification retained. Squares require at least 80% photographic support and 80% classification within photographed area. Technical definition: 100 W/K; W is woody-labelled area and K is accepted classified area. Unknown and missing areas are excluded; this is not whole-cell tree cover.
2020 lidar: all nonwithheld, nonoverlap, nonnoise first returns; ground normalized with class-2 terrain TIN and 15 m neighbor buffer. A return is supported only within a triangle, with nearest ground ≤3 m and triangle edges ≤10 m. Height is p99 of all supported first returns. Fraction is the percentage of supported first returns ≥2 m. Roofs may contribute; negative heights are retained. Bounds −1 to 60 m are diagnostic flags, not clipping limits or confidence intervals. Unsupported returns are excluded, never forced to zero.
Ground terrain: USGS 3DEP 10 m DEM, source acquisition range 2020-05-26 to 2021-03-13. Central differences in projected metres; invalid neighbor/edge derivatives masked. Display slope is area-weighted native slope (degrees). Aspect is downhill clockwise from north using the area-weighted gradient; flat or cancelling gradient is undefined. Not a photo-minus-DEM height calculation.
LANDFIRE 2024 product year, Scott/Burgan40 context: native 30 m EPSG:5070 grid. Display uses the largest intersecting category by projected area; mixed-cell proportions are preserved on click and export. IDs are never averaged. Native export preserves original source pixels and registration. Short model codes and group names are verified from the locally installed pyretechnics table; unknown IDs stay IDs. Colours are categorical, not a fuel-loading or hazard scale.
No temporal-change inference. No 2021 parity claim. First implement tree/non-tree exclusion, crown delineation, minimum crown area and forest rules; base height and bulk density additionally require exact model inputs and coefficients.