# N05_E18 Great Lakes pilot: authoritative WBD geometry and fixed-cell intersection

**Run date:** 2026-08-11  
**Sector:** `N05_E18`  
**Canonical geometry:** latitude 37.5–45.0° N; longitude 270–285° E, equivalent to 90–75° W  
**Time bands:** modern baseline (2000–present), with the hydrologic geometry used as a present-day anchor for historical crosswalks  
**Research status:** bounded finding; no quadrant or sector-wide score assigned

## Selection and non-duplication decision

This run continues the Great Lakes pilot and follows the explicit next path in the project status notes: retrieve the full USGS Watershed Boundary Dataset (WBD) layer-5 geometry for HUC-10 `0404000105` and calculate its intersection with the fixed `N05_E18` cell. This is a geometry/provenance run, not a new Gary industrial-history extraction.

An Archive.org search for additional Lake County/Gary waterworks and industrial records was also performed. The most relevant result, **Gary's Golden Jubilee 1906–1956**, is already documented in the existing 1956 Gary Works notes; its OCR and page-level industrial/waterway observations were not duplicated here. The archival search therefore produced no new defensible historical observation for this run.

## Spatial definition and method

The fixed sector is represented by the canonical rectangular envelope from the project sector record:

```text
latitude:  37.5 to 45.0 degrees north
longitude: 270 to 285 degrees east = -90 to -75 degrees west
```

The WBD feature was requested from the official USGS National Map REST service with `HUC10='0404000105'`, all attributes, polygon geometry, and EPSG:4326 output. Geometry was checked with Shapely, then transformed from EPSG:4326 to EPSG:5070 (NAD83 / Conus Albers) before area calculations. The intersection was calculated between the returned polygon and the fixed rectangular envelope. EPSG:5070 was used for area rather than degrees because degree-area arithmetic is not an appropriate comparison across this latitude band.

The method is reproducible from the request URL and the project sector bounds. It does not clip the cell to land, shoreline, or a political boundary; the result is the area share of the hydrologic object within the canonical rectangular sector envelope.

## Raw observations

### A. USGS WBD feature response

The official WBD layer-5 response returns one polygon for HUC-10 `0404000105`:

| Field | Direct observation | Provenance | Confidence | Limitation |
|---|---|---|---|---|
| `huc10` | `0404000105` | USGS WBD layer-5 query | high | Current dataset object; not a historical boundary |
| `name` | **Little Calumet River-Deep River** | USGS WBD layer-5 query | high | Name applies to the HUC-10 object, not the full sector |
| `areasqkm` | `507.98 km²` | WBD attribute response | high | Dataset-reported area; not independently substituted for the geometry calculation |
| `sourceoriginator` | Natural Resources and Conservation Service and U.S. Geological Survey | WBD attribute response | high | Source metadata, not a claim about historical stewardship |
| `loaddate` | `2024-08-15T13:01:21Z` | WBD attribute response, epoch milliseconds `1723726881000` | high | Service load date, not the observation date of every landscape feature |
| geometry | One valid polygon; longitude bounds approximately `-87.4953224` to `-87.1519895`, latitude bounds approximately `41.3621476` to `41.6340396` | Returned GeoJSON geometry | high | Bounds describe the returned HUC-10 polygon only |
| `hutype` | `S` | WBD attribute response | high | Code retained as reported; not interpreted as a cultural or institutional category |
| `humod` | `NM` | WBD attribute response | high | Code retained as reported; code semantics require the service schema if used later |

### B. Archive.org search audit

The Archive.org search returned **Gary's Golden Jubilee 1906–1956**, an item already used by the project. Its metadata identifies the creator as the Gary Golden Jubilee Committee, date 1956, a 132-page publication, and Lake County Public Library as submitting institution. Existing notes already record page-checked observations on Gary Works, Grand Calumet channel modification, harbor construction, water infrastructure, rail tracks, and industrial scale. Because this run’s selected unresolved path was the WBD/cell geometry, those existing observations were not re-extracted or counted as new findings.

This is a search-status observation, not evidence that no other historical source exists:

```text
archive_search_status: relevant duplicate source located; no new archival observation retained
missingness: historical_measure_not_available for a new source in this run
```

## Derived geometry

The following values are derived from the raw WBD polygon and the fixed sector envelope. They are not source-reported values.

| Derived quantity | Result | Formula / method | Confidence | Uncertainty |
|---|---:|---|---|---|
| WBD polygon area | `507.984365 km²` | EPSG:5070 polygon area | high | Small difference from the attribute-rounded `507.98 km²`; geometry/service version controls this calculation |
| Fixed rectangular sector-envelope area | `1,042,110.404 km²` | EPSG:5070 area of `[-90,-75] × [37.5,45]` | medium-high | This is the canonical rectangular envelope, not a land-only or shoreline-clipped area |
| HUC-10 ∩ `N05_E18` area | `507.984365 km²` | EPSG:5070 area of polygon intersection | high | Depends on the current WBD geometry and the fixed rectangular cell definition |
| HUC-10 area retained inside the cell | `100.000000%` | intersection area ÷ WBD polygon area × 100 | high | Geometric containment at the current service resolution; does not imply historical continuity |
| HUC-10 share of fixed sector envelope | `0.0487457%` | intersection area ÷ sector-envelope area × 100 | medium-high | The denominator includes the full rectangular cell, including areas outside the local watershed and mixed land/water portions |

The returned polygon is entirely within the canonical sector envelope at the requested geometry. Therefore, for this local anchor, the previous `geographic_match_uncertain` status can be narrowed: the **current WBD object’s intersection with the canonical rectangular envelope is resolved**, while cell-wide hydrologic composition remains unresolved.

## Source interpretation

The modern WBD record establishes a reproducible local hydrologic object inside `N05_E18`: the Merrillville anchor’s HUC-10 lies wholly within the canonical sector envelope and occupies approximately 507.98 km², or 0.04875% of the full rectangular cell envelope. This is useful for linking point, reach, and local-basin observations to a named drainage object without pretending that the drainage object is the sector.

The result does **not** establish that the entire sector drains to Little Calumet River–Deep River. It also does not establish that the 1921 Lake County transport, drainage, or industrial descriptions map continuously onto the present WBD object. The WBD is a modern hydrologic classification with a 2024 load date; it is not a historical reconstruction of drainage boundaries or land use.

The small area share is a geometry result, not a measure of hydrologic importance, population, industrial intensity, or water availability. A small watershed can still be a locally important transport, settlement, or contamination pathway. No variable was normalized into a sector score.

## Variable crosswalk

| Historical or modern variable | WBD geometry result | Comparability | Boundary note |
|---|---|---|---|
| 1921 county drainage/land-conversion narrative | Modern HUC-10 local anchor is now geometrically located inside `N05_E18` | related but not equivalent | County narrative and modern watershed have different dates and footprints |
| Historical river/canal/port descriptions | HUC-10 polygon provides a present-day drainage context | related but not equivalent | WBD does not prove historical channel continuity or port function |
| Modern watershed area | `507.984365 km²` geometry-derived area | directly comparable only to the same WBD object/version | Not comparable to fixed-cell area without retaining the denominator and CRS |
| Fixed-cell water composition | Not established by one local HUC-10 | not comparable in this run | Additional hydrologic units, shoreline/lake geometry, and the full cell boundary must be assembled |

## Derived constraints and adaptive-capacity indicators

No model formula was applied. The geometry supports only these bounded candidate inputs:

- **Hydrologic localization candidate:** HUC-10 `0404000105` is a reproducibly identified local drainage object wholly inside the fixed cell.
- **Hydrologic footprint candidate:** the object’s current mapped area is 507.984365 km², with the source-reported rounded area retained as 507.98 km².
- **Boundary sensitivity candidate:** a local watershed is much smaller than the rectangular cell envelope; future sector aggregation must not silently substitute watershed statistics for cell-wide values.

`provisional_quadrant`: **unresolved**. The run does not supply the comparable raw variables, formula version, or historical coverage required by the SOP for classification.

## Missingness and uncertainty

- `historical_measure_not_available`: no historical drainage polygon or 1921-to-present channel-continuity series was established.
- `geographic_match_uncertain`: resolved for the returned HUC-10’s containment within the canonical rectangular envelope; unresolved for the relationship between the HUC-10 and the full sector’s hydrologic composition.
- `period_match_uncertain`: WBD geometry is a modern snapshot with a 2024 load date; it is not a 1800–1950 or 1921 boundary record.
- `fixed_cell_water_composition_missing`: no complete cell-wide lake, river, watershed, shoreline, or coastal-water inventory was derived.
- `source_exists_not_accessible`: no new primary administrative or engineering record was retained from the Archive.org search audit.
- `not_applicable`: no quadrant and no derived constraint score.

Archival silence, the absence of a new source in this search, and the small HUC-10 share of the cell were not converted into zeros or negative evidence.

## Falsification and alternative explanations

The result weakens any interpretation that treats one local watershed as a proxy for the whole sector. If a later model changes materially when it uses the fixed cell, watershed, industrial corridor, or labor-market boundary, that difference must remain visible. The observed overlap could also reflect the choice of Merrillville as a human anchor rather than a sector-wide hydrologic pattern.

Alternative explanations for historical industrial concentration remain open: corporate site selection, lake shipping, rail investment, labor migration, municipal policy, land acquisition, state and federal infrastructure, and legal power may explain more of the observed pattern than hydrology alone.

## Next search path

1. Use the resolved HUC-10 polygon as one local anchor while retrieving the remaining WBD/NHD or NLDI objects that intersect the canonical `N05_E18` envelope; preserve each object’s identifier, geometry version, and overlap rather than collapsing them into one watershed value.
2. Retrieve a machine-readable shoreline/lake and stream-network layer to distinguish hydrologic units from Lake Michigan and other water surfaces inside the mixed land/water cell.
3. Crosswalk the HUC-10 and NLDI objects to the historical 1921 drainage, port, canal, and municipal-water observations already documented in the project; do not infer continuity from spatial proximity alone.
4. Continue the 1950–2000 bridge and complete comparable raw-observation tables for the nine Great Lakes sectors before assigning any quadrant.

## Provenance register

| Source | Tier | Item / page / endpoint | Date accessed | URL |
|---|---|---|---|---|
| USGS Watershed Boundary Dataset service | A: authoritative machine-readable dataset | National Map WBD service metadata and layer structure | 2026-08-11 | <https://hydro.nationalmap.gov/arcgis/rest/services/wbd/MapServer> |
| USGS WBD layer-5 query | A: authoritative machine-readable response | `where=HUC10='0404000105'`, `outFields=*`, `returnGeometry=true`, `outSR=4326`, GeoJSON response | 2026-08-11 | <https://hydro.nationalmap.gov/arcgis/rest/services/wbd/MapServer/5/query> |
| USGS WBD documentation | A: institutional technical documentation | Definition, nested hydrologic-unit system, and dataset status | 2026-08-11 | <https://www.usgs.gov/national-hydrography/watershed-boundary-dataset> |
| Canonical project sector record | A: project record | `N05_E18`, 37.5–45.0° N, 270–285° E, Merrillville anchor | 2026-08-11 | `N05_E18_MERRILLVILLE-sector-record.json` |
| Gary Golden Jubilee 1906–1956 | C: digitized local historical publication; duplicate-search audit only | Internet Archive item metadata and OCR derivative; existing project notes contain the page-level extraction | 2026-08-11 | <https://archive.org/details/garysgoldenjubilee19061956> |
| Gary Golden Jubilee OCR derivative | C: locator/transcription aid; duplicate-search audit only | Item OCR derivative | 2026-08-11 | <https://archive.org/download/garysgoldenjubilee19061956/Gary%27s%20Golden%20Jubilee%201906-1956_djvu.txt> |
| Project SOP | A: project procedure | Fixed 24 × 24 geometry, source hierarchy, provenance, missingness, and classification boundary | 2026-08-11 | `08-research-sop.md` |

## Bottom line

The full current WBD geometry for HUC-10 `0404000105` is now retrieved and audited. The polygon is valid, lies entirely inside the canonical `N05_E18` rectangular envelope, and has an equal-area geometry-derived area of `507.984365 km²` versus the WBD attribute’s rounded `507.98 km²`. The HUC-10 therefore contributes `100%` of its own area to the cell and approximately `0.0487457%` of the full rectangular cell envelope.

This resolves the local HUC-10/fixed-cell intersection requested by the prior research path. It does not resolve the cell’s full hydrologic composition, historical drainage continuity, comparable historical/modern variables, the formula, or `provisional_quadrant`, which remains **unresolved**.