California History Maps

Lumber Ports of the California Coast · Sources & Method

Companion to the interactive map. Last updated 2026-09-13. Every pin on the map carries its own citations in its popup; this page states how the dataset was built, what its classifications rest on, and where its evidence is uneven.

1. The backbone inventory, and its draft status

The Sonoma–Mendocino core of this map follows the 57-port inventory of Deborah Marx & Denise Jaffke, Northern California Doghole Ports Maritime Cultural Landscape, National Register of Historic Places Multiple Property Documentation Form (California Dept. of Parks and Recreation / OHP, draft, rev. 2021). Because it is a draft nomination, its inventory numbers and inclusion decisions could change before listing. Several small-port entries here rest on sources quoted within the MPDF (Davidson page citations, Jackson 1977, local newspapers) that were not independently re-opened; popups mark these "via the MPDF." Claims read directly from a primary source carry a [P] tag.

2. Dating rule: the Coast Pilot lag

Activity dates are ranges, not points, and many are keyed to editions of the U.S. Coast Pilot: George Davidson's Coast Pilot of California, Oregon, and Washington (1869; 4th ed. 1889). An edition reflects field reconnaissance several years before its imprint (the 1889 edition largely describes c.1885–87), so "abandoned by 1889" means absent from that survey, not a closure date. Ports founded after ~1887 are structurally invisible in Davidson; their dates come from the MPDF, newspapers, and local records. Absence from Davidson is not evidence of absence. All dates on this map labeled circa should be read with that lag in mind.

3. The four-tier port classification and its warrants

A classification resting on a single T-sheet or one adjectival phrase carries lower confidence; ports with no surviving structure description are classed inferentially. Large steam schooners did occasionally serve wire-chute dogholes (the Scotia at Fish Rock, 1914, per the MPDF), so the tiers are evidence-based generalizations, not laws.

4. Coordinates: provenance and validation

Every pinned feature's popup states its coordinate source: a USGS GNIS record (with feature ID), Coast Pilot verbal geometry ("mouth of Hare Creek," with the GNIS stream-mouth position), an NRHP/archaeological report, or a published structure position. Precision codes: exact (source pins the structure), place (named cove, point, or creek mouth; expect ~100–300 m), area (right stretch of coast, ~400 m+), conjectural. The merged set was validated by runnable gates: north-to-south monotonic latitude down the canonical 57 (zero violations), Davidson's own stated inter-port distances (7/7 within tolerance), and coastal-band checks. Six documented ports have no confirmed location and are deliberately not pinned: Monroe, Devilbliss, the Unnamed Landing near Uncle Abe's, Uncle Abe's, Buster's, and Scott's (the last two share one chart candidate and may be a single site). Chute-level refinement from georeferenced Coast Survey T-sheets (NOAA Shoreline Data Explorer) is planned and pending.

5. Railroad alignments

Rail lines are drawn along their actual corridors where geometry exists: the California Western's mapped, still-operating line (OpenStreetMap; 63 km stitched, matching the historic 40 miles); the Northwestern Pacific, Arcata & Mad River, and Scotia–Alton corridors likewise. Vanished logging grades are drawn along the named creek valleys their sources record (USGS National Hydrography Dataset flowlines) and render dashed as reconstructed corridors; a dashed line means "the grade ran up this valley," never a surveyed alignment. Each route's popup cites its historical sources and its geometry source separately. The Bear Harbor & Eel River line's fixed points (pier, gravity incline, Moody, Andersonia) come from the Bear Harbor Landing NRHP nomination; its creek-corridor densification snapped to within meters of the documented incline head and Moody townsite.

6. Coverage asymmetry

The evidence base is not uniform, and uniform symbology can mask that. Sonoma and Mendocino counties inherit a dense professional apparatus (the MPDF, NOAA/State Parks doghole archaeology, Fort Ross research design, Palmer 1880). Humboldt and Del Norte rest on county histories, NPS studies, and repository collections. Marin, San Mateo, Santa Cruz, Big Sur, and San Luis Obispo features were assembled from thinner published bases and, following a 2026-09-13 source audit, carry added primary and scholarly citations (Davidson 1889; Stanger; Jepson's tanbark bulletin; UC Press regional histories); but no inventory of MPDF rigor exists south of the Golden Gate. Superlatives ("the highest and longest wire chute") remain attributed to their period sources, never flattened to fact.

7. Photographs

59 features carry a historic image (one deliberately modern and labeled as such: never-used Bromley Gulch). Images stored on this site are U.S.-government works (NOAA, HAER/HABS, USCG), pre-1930 U.S. publications, or repository scans of pre-1930 photographs in the public domain; each carries a printed credit (repository, collection, photographer, date). A few images are hot-linked from their repositories with credit rather than copied. Images from museums and hobby sites that are not rights-cleared are cited as leads in the research files, not embedded. One identification (Bear Harbor) is marked tentative in its credit line, following its catalog record.

8. The Mexican-period prelude

Reed's Sawmill (Rancho Corte Madera del Presidio) is included with its construction date presented as genuinely disputed; "about 1833-34" in the landmark tradition (CHL No. 207) against 1843 in Munro-Fraser's detailed 1880 account; and with the rancho-name tradition ("cut wood for the Presidio") assigned to the pre-grant woodcutting trade rather than the mill. No Mexican-period landing for the mill is attested; both rancho diseños show no molino.

9. Principal sources

Primary

Secondary

Archives & collections

10. Corrections and the claims audit

On 2026-09-13 the full dataset underwent an adversarial claims audit: ~382 discrete factual claims (dates, operators, structures, tonnages, superlatives) were checked against their sources; the MPDF and both Davidson Coast Pilots read directly, HAER data pages, county histories, and every cited web source; with a verdict recorded for each. About two-thirds were confirmed outright; 68 were corrected in place (examples: Weller & Stewart operated Bear Harbor 1889-92, not 1890-92; Samoa's 1956 buyer was Georgia-Pacific, not Louisiana-Pacific; the Albion logging line began as 40-inch narrow gauge; the Año Nuevo wharf was gone by 1883 despite Davidson's carried-over 1889 description) and 54 unsupported claims were softened, attributed, or removed (unsourced superlatives; company attributions the record does not bear). Every photograph was re-verified against its catalog record, and one image NOAA captions inconsistently (Fisk Mill Cove vs Salt Point) now appears once, with the conflict disclosed. Claims the audit could not resolve are flagged "unverified" in their popups rather than silently retained. Found something wrong? See the site's errata page.