The archive's own cartography · data, not pictures

GIS package

Every mapped layer the archive holds and may redistribute, plus the knowledge base itself as geometry, in the formats a GIS opens. Load it on top of whatever basemap you already use — 36 vector layers, 22,218 features, in EPSG:4326, plus the island's bare-earth LiDAR at its full one-metre resolution. Free to download and use under the terms each layer carries.

The part that exists nowhere else is archive_points, archive_lines and archive_polygons: 487 records from this archive, each carrying a link back to its article. Click a point in QGIS and you land on the sources.

Downloads

File Geodatabase6.7 MB · beaver-island-archive.gdb.zip
ArcMap and ArcGIS Pro
For Esri users whose version will not read a GeoPackage. Unzip first. Layers with mixed geometry are split, so osm_water arrives as osm_water_poly and osm_water_line.
GeoPackage28.9 MB · beaver-island-archive.gpkg
QGIS, ArcGIS Pro, R, Python — anything that speaks OGC
Every vector layer in the package, styles built in: add a layer and it draws in the archive's own colours. Carries a bia_layer_metadata table naming every layer's source and licence.
Google Earth207 kB · beaver-island-archive.kmz
Google Earth, desktop or web
The archive's own records plus roads, trails, names and historic sites. Click a placemark for a link straight to its article. Deliberately not the whole package — 39,000 parcel and soil polygons make a KMZ unusable.
QGIS project4 kB · beaver-island-archive.qgz
QGIS 3.x and later
Opens the GeoPackage with the layers grouped by theme over an OpenStreetMap basemap. Keep it beside the .gpkg — it looks for it in the same folder.

Also published: layers.csv — the full layer table with source and licence for each — and manifest.json.

The terrain

Beaver Island was flown with airborne LiDAR in 2018 at one metre, bare earth — the vegetation and buildings stripped out, leaving the ground. It is the layer that shows what is still there under sixty years of second-growth: the old road grades, the beach ridges the lake left behind as it fell, the rectangular dip where a cellar was.

If you only want to see the ground, take archive_contours_2m from the GeoPackage above — 16,235 contour lines at a two-metre interval, drawn from this same LiDAR. Vectors draw over any basemap, label themselves and can be queried, and the whole layer is a few megabytes rather than a few hundred.

Bare-earth elevation model — 1 m, full resolution460.1 MB · beaver-island-dem-1m.tif
QGIS, ArcGIS, GDAL — a Cloud Optimized GeoTIFF
The LiDAR itself at its own resolution: 12,765 x 22,340 cells of ground elevation in NAVD88 metres, vegetation and buildings already removed. This is the file to measure from — a road prism, a beach ridge, a cellar hole. Being a COG, a client that only needs a corner fetches only that corner, and QGIS draws it from the built-in overviews without reading all 460 MB.
Bare-earth elevation model — 2 m122.3 MB · beaver-island-dem-2m.tif
The same thing, a quarter the size
Resampled to 2 m for people who want the shape of the island rather than a measurement. It still resolves the beach ridges and road prisms; it is not what to cite a height difference from.
Hillshade — 1 m106.1 MB · beaver-island-hillshade-1m.tif
A drop-in backdrop for any map
The same LiDAR rendered as shadow-only illumination, 8-bit, in the archive's own light. Put it under the vector layers and everything sits on real ground.
Hillshade — 2 m26.4 MB · beaver-island-hillshade-2m.tif
The lighter backdrop
Identical rendering at half the resolution. Below about 1:5,000 you will not see the difference.

All four are Cloud Optimized GeoTIFFs in EPSG:3078 (Michigan GeoRef), with overviews built in, so a viewer draws them at a sensible zoom without reading the whole file. Elevations are metres above the NAVD88 geoid, not feet and not depth below the lake. Source: USGS 3DEP, Charlevoix County 2018, QL2 — public domain.

Loading it

QGIS: Download the GeoPackage and the project file into the same folder, then open the .qgz. The layers arrive grouped by theme over an OpenStreetMap basemap, already styled — the styles live inside the GeoPackage, so dragging in a single layer on its own works too. The terrain rasters are deliberately not in the project file, since most people want one of the four rather than all of them: drag a .tif in and it georeferences itself into place.
ArcGIS Pro: Pro reads a GeoPackage directly: Add Data, then pick the .gpkg. Symbology is QGIS-specific and will not carry over, so the layers arrive in Pro's default colours.
ArcMap, or any Esri version that will not read a GeoPackage: Use the File Geodatabase. Unzip it first and add the .gdb folder. Layers holding more than one geometry type are split — osm_water arrives as osm_water_poly and osm_water_line — because a geodatabase permits one geometry type per feature class.
Google Earth: Open the KMZ. It carries the archive's records plus roads, trails, names and historic sites, and each placemark links to its article. It is deliberately not the whole package: 39,000 parcel and soil polygons make a KMZ unusable.
The terrain rasters: A Cloud Optimized GeoTIFF is an ordinary GeoTIFF — QGIS, ArcGIS Pro, gdalinfo, rasterio and terra open it with no special handling, and nothing needs unzipping. Take the 1 m files if you are measuring anything; the 2 m files are there for when the download matters more than the metre.
R, Python, anything else: A GeoPackage is a SQLite database. sf::st_read(), geopandas.read_file() and ogr2ogr all read it, and bia_layer_metadata inside it names every layer's source and licence.

What is in it, and what is not

The archive holds more layers than it may redistribute. This package carries only those its publication ledger allows, which is why some layers you can see on atlas plates are not here — a licence the classifier could not recognise is excluded rather than shipped on an assumption. That is a statement about review, not about restriction.

Parcel owner names and owner mailing addresses removed. Dataset-catalogue records excluded; their source layers ship whole.

Every layer's licence and source travel with it, in the layer table above and in a bia_layer_metadata table inside the GeoPackage, so the data cannot be separated from its terms. OpenStreetMap layers are © OpenStreetMap contributors under ODbL 1.0. The archive's own layers are CC BY 4.0.

The 36 layers

The archive's own records 4

Layer	Features	Geometry	Licence
`archive_points`	348	Point	CC BY 4.0 — Beaver Island Archive; underlying sources vary, see each record
`archive_lines`	135	LineString, MultiLineString	CC BY 4.0 — Beaver Island Archive; underlying sources vary, see each record
`archive_polygons`	4	Polygon	CC BY 4.0 — Beaver Island Archive; underlying sources vary, see each record
`archive_contours_2m`	16,235	LineString	USGS 3DEP (public domain); contours CC BY 4.0 Beaver Island Archive

Land, water and boundaries 6

Layer	Features	Geometry	Licence
`federal_gnis_gnis_names`	82	Point	Public domain (U.S. Geological Survey, GNIS)
`integrated_gazetteer`	242	Point	Mixed — see contributing source provenance; OSM portions ODbL
`osm_islands`	17	MultiPolygon, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_landuse`	57	MultiPolygon, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_natural`	439	LineString, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_water`	225	LineString, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors

Roads, trails and routes 3

Layer	Features	Geometry	Licence
`osm_aeroway`	37	LineString, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_roads`	414	LineString	ODbL 1.0 — © OpenStreetMap contributors
`osm_trails`	139	LineString	ODbL 1.0 — © OpenStreetMap contributors

Buildings, places and structures 7

Layer	Features	Geometry	Licence
`osm_amenities`	88	LineString, MultiPolygon, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_buildings`	955	Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_cemeteries`	3	Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_historic`	14	LineString, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_lighthouses_maritime`	64	LineString, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors
`osm_places`	5	Point	ODbL 1.0 — © OpenStreetMap contributors
`osm_power_utilities`	8	LineString, Point, Polygon	ODbL 1.0 — © OpenStreetMap contributors

Public Land Survey and parcels 3

Layer	Features	Geometry	Licence
`federal_plss_plss_intersected`	1,284	Polygon	Public domain (U.S. Bureau of Land Management, CadNSDI)
`federal_plss_plss_section`	147	MultiPolygon, Polygon	Public domain (U.S. Bureau of Land Management, CadNSDI)
`federal_plss_plss_township`	16	MultiPolygon, Polygon	Public domain (U.S. Bureau of Land Management, CadNSDI)

Geology and landforms 2

Layer	Features	Geometry	Licence
`geology_usgs_sgmc_contacts`	8	LineString, MultiLineString	Public domain (U.S. Geological Survey)
`geology_usgs_sgmc_lithology`	8	MultiPolygon, Polygon	Public domain (U.S. Geological Survey)

Hydrography and wetlands 3

Layer	Features	Geometry	Licence
`federal_nhd_flowline`	57	LineString	Public domain (U.S. federal government)
`federal_nhd_waterbody`	27	Polygon	Public domain (U.S. federal government)
`federal_nwi_wetlands`	496	Polygon	Public domain (U.S. Fish and Wildlife Service)

Soils, land cover and forest 2

Layer	Features	Geometry	Licence
`soils_ssurgo_mapunit_polygons`	553	Polygon	Public domain (USDA Natural Resources Conservation Service)
`soils_ssurgo_mapunit_polygons_mi089`	56	Polygon	Public domain (USDA Natural Resources Conservation Service)

Protected land and heritage 4

Layer	Features	Geometry	Licence
`federal_nrhp_points`	2	Point	Public domain (National Park Service)
`federal_nrhp_polygons`	1	Polygon	Public domain (National Park Service)
`federal_padus_fee_managers`	20	MultiPolygon, Polygon	Public domain (USGS Protected Areas Database)
`osm_protected_areas`	20	MultiPolygon, Polygon	ODbL 1.0 — © OpenStreetMap contributors

Administrative and demographic 2

Layer	Features	Geometry	Licence
`demographics_acs_tracts`	6	Polygon	Public domain (U.S. Census Bureau)
`osm_boundaries`	6	MultiPolygon, Polygon	ODbL 1.0 — © OpenStreetMap contributors