public-connectivity

Landscape connectivity & climate-resilience datasets. Most cover California, from the California Biodiversity Network (CBN) and partners; rcn-national is continental, covering the conterminous US and adjacent Canada. All products are cloud-native and served from NRP S3 (https://s3-west.nrp-nautilus.io/public-connectivity/).

All layers are in OGC:CRS84 / EPSG:4326. The California raster sources were reprojected from their native ESRI:102003 (USA Contiguous Albers) / sphere Web-Mercator projections to EPSG:4326 (categorical layers with nearest-neighbor; the continuous flow index with bilinear).

---

Vector — regional-connectivity-linkages (Beier 2006)

11 South Coast Missing Linkages polygons (e.g. Tehachapi, Sierra Madre – Castaic).

DuckDB:

MapLibre GL JS (PMTiles — note the source-layer is regional-connectivity-linkages):

H3 hex (res 8). Area columns are per-feature totals repeated on every cell — never SUM them on hex; dedup by _cng_fid, or compute area from COUNT(DISTINCT h8) × cell_area_at_res_8:

---

Vector — wildlife-movement-barriers (CDFW 2024)

204 linear infrastructure segments (roads/highways) that CDFW identified as barriers to terrestrial wildlife movement across California, with priority rankings for 2020/2022/2024 and remediation status. Source: CDFW BIOS ds2867 (priorities as of June 2024). Geometry is lines (LineString / MultiLineString); each segment was hexed to H3 resolution 8 by buffering it by the H3 cell circumradius before polyfill.

DuckDB — top-priority barriers by length:

MapLibre GL JS (PMTiles — note the source-layer is wildlife-movement-barriers):

H3 hex (res 8). Length_kilometers is a per-feature total repeated on every cell a barrier covers — never SUM it on hex; dedup by Barrier_ID (one row per barrier) or _cng_fid first, or count barriers with COUNT(DISTINCT Barrier_ID):

Priority_2020/Priority_2022/Priority_2024 take values Barrier, Priority Barrier, Top Priority Barrier, Remediated (NULL = not assessed that cycle). Target_Species is a free-text, comma-separated list of focal species. CDFW_Region is 1–6 (comma-joined for barriers spanning multiple regions). SUMMARY_LINK is a per-barrier CDFW summary PDF.

---

Rasters

H3 hex tables have columns <value>, h9 (native res 9), h0 (partition key). One row per H3 cell.

climate-migration-routes (Schloss 2022) — categorical, 13 classes

Connectivity class per cell (mode reducer). Classes: 0 limited/none; 100–103 Diffuse (intact); 200–203 Intensified (multiple linkages); 300–303 Channelized (present-day linkage); +1/+2/+3 = HADGEM2-ES / CNRM-CM5 / both climate-model linkages. See classification:classes on the COG band.

COG (titiler / categorical colormap): https://s3-west.nrp-nautilus.io/public-connectivity/climate-migration-routes-cog.tif

present-day-connectivity/flow (Cameron 2022) — continuous index 0–100

Current-flow connectivity index (mean reducer).

COG: https://s3-west.nrp-nautilus.io/public-connectivity/present-day-connectivity-flow-cog.tif

present-day-connectivity/categories (Cameron 2022) — categorical, 12 classes

Connectivity-strategy category per cell (mode reducer). Classes: 3 Limited movement potential; 4 Permeable (local movement); 19 Impeded; 25,29 Diffuse; 31,35,39 Intensified; 41,45,49 Channelized; 0 background (undocumented in source). See classification:classes on the COG band. COG: https://s3-west.nrp-nautilus.io/public-connectivity/present-day-connectivity-categories-cog.tif

---

Citations

• California Department of Fish and Wildlife (2024). Terrestrial Wildlife Connectivity Barriers (Wildlife Movement Barriers). CDFW BIOS ds2867. https://wildlife.ca.gov/Data/BIOS
• Beier, P., et al. (2006). South Coast Missing Linkages. South Coast Wildlands. (CDFW BIOS ds419)
• Cameron, D. R., Schloss, C. A., Theobald, D. M., & Morrison, S. A. (2022). A framework to select strategies for conserving and restoring habitat connectivity. Conservation Science and Practice, e12698. https://doi.org/10.1111/csp2.12698
• Schloss, C. A., et al. (2022). "No-regrets" pathways for navigating climate change. Ecological Applications. https://doi.org/10.1002/eap.2468

---

Raster - rcn-national (TNC Resilient and Connected Network 2024)

The Nature Conservancy's Resilient and Connected Network at 30 m: the sites and corridors that together would sustain biodiversity under a changing climate, classified into 14 classes. Footprint is the conterminous US plus adjacent Canada (the prairie provinces north to about 52.7 degrees, and New Brunswick, Nova Scotia and Prince Edward Island), so a US statistic needs a national mask. Alaska and Hawaii are separate source rasters and are not included.

rcn_class values: 0=Not in Network, 1=Resilient Not in Network, 10=Resilient, Diffuse Flow (Climate Informed), Recognized Biodiversity Value, 20=Resilient, Diffuse Flow (Climate Informed), 30=Resilient, Diffuse Flow, Recognized Biodiversity, 40=Resilient, Diffuse Flow, 50=Resilient, Concentrated Flow (Climate Informed), Recognized Biodiversity, 60=Resilient, Concentrated Flow (Climate Informed), 70=Mostly Resilient, Concentrated Flow, Recognized Biodiversity, 80=Mostly Resilient, Concentrated Flow, 90=Resilient Coastal Migration Space, 91=Resilient, Recognized Biodiversity, Secured, 92=Additional Resilient Secured (GAP 3), 93=Resilient, Recognized Biodiversity. 0 and 1 are real classes (mapped land outside the network, together about 62% of the footprint) - only 255 is no data.

Licence: TNC non-commercial. That restriction carries to anyone using these products.

COG (titiler, categorical colormap from the STAC classification:classes):

https://s3-west.nrp-nautilus.io/public-connectivity/rcn-national-cog.tif

H3 hex (native resolution 10, mode reducer) - dominant class per cell:

H3 hex, fractional coverage (one row per cell and class present) - use this, not the mode layer, for class areas; weight each frac by its own cell's ground area (see the h3-guide) rather than counting cells:

There is no PMTiles or GeoParquet for this dataset: raster ingests produce a COG and hex only.