# N05_E18 Great Lakes pilot: 1956 Gary Works anchor and 2021–2024 WQP profile **Run date:** 2026-08-06 (America/Chicago) **Sector:** `N05_E18` **Canonical geometry:** 24 latitude bands × 24 longitude bands = 576 sectors; this note uses the existing N05_E18 cell and does not redefine it. **Time band:** 1950–2000 bridge, with a modern structured-data endpoint extending to 2024. **Status:** bounded research result; no sector score or quadrant assigned. ## Scope and selection This run continues the Great Lakes pilot and does not expand globally. It follows the preceding N05_E18 notes’ next-search path: extract the analyte-level WQP profile for `USGS-05536356`, then add a distinct Archive.org historical anchor for Gary’s industrial scale in the 1950s. The two sources are not treated as equivalent measurements. The WQP file is a station-level observation set; the Gary source is a contemporary local historical publication describing city and plant history. The fixed cell is larger than either the station point or Gary municipal footprint. The historical source’s industrial claims concern Gary and the Calumet industrial corridor, which cross the cell and do not automatically describe every part of `N05_E18`. ## Research question What provenance-bearing observations can connect the mid-century industrial context to a modern machine-readable water-quality record without converting narrative scale claims into cell-wide measurements or assigning a model quadrant? ## Raw observations ### A. Water Quality Portal station metadata The WQP station CSV identifies `USGS-05536356` as **GRAND CALUMET RIVER AT COLUMBIA AV AT HAMMOND, IN**, a stream station in Lake County, Indiana, with latitude `41.6186111`, longitude `-87.4998333`, HUC8 `07120003`, and provider `NWIS`. The station’s source-map scale is 1:24,000; horizontal collection is reported as interpolated from a digital map; horizontal datum is NAD83. These are station metadata, not a sector-average water condition. **Source record:** WQP station CSV, one returned station row, retrieved 2026-08-06. **Source URL:** ### B. Machine-readable result coverage The WQP result CSV for `USGS-05536356` returned **849 water rows**, covering **40 distinct sample dates from 2021-05-24 through 2024-02-22**. The rows contain **33 distinct characteristic names**. The source reports 624 rows as `Preliminary` and 225 as `Accepted`; 12 rows have `ResultDetectionConditionText = Not Detected`. All returned rows are `ActivityMediaName = Water`. The file contains observations from multiple dates, methods, fractions, units, and status classes. These rows are preserved as reported results rather than collapsed into an unapproved period mean. ### C. Selected reported result fields The following are raw source observations from the returned CSV. Values are not interpreted as a water-quality score. | Characteristic | Date range in returned profile | Reported units | Descriptive range across numeric rows | Status / qualification note | |---|---:|---|---:|---| | `Temperature, water` | 2022-06-06 to 2022-12-07 in sampled rows | `deg C` | 4.4–27.8 | 39 numeric rows; mix of `Accepted` and `Preliminary` | | `Oxygen` | 2022-06-06 to 2022-12-07 in sampled rows | `mg/l` and `% saturatn` | 5.3–674.0 within the mixed-unit field | Units must not be combined; the 674 mg/l row is preliminary and requires source-method review | | `pH` | 2022-06-06 to 2022-12-07 in sampled rows | `std units` | 7.1–8.1 | 39 numeric rows; mix of `Accepted` and `Preliminary` | | `Specific conductance` | 2022-06-06 to 2022-12-07 in sampled rows | `uS/cm @25C` | 819–1,630 | 51 numeric rows; duplicate-looking rows retained pending activity/method review | | `Phosphorus` | 2022-06-06 to 2022-12-07 in sampled rows | `mg/l as P` | 0.07–0.72 | 30 numeric rows; status and method fields preserved | These ranges are descriptive checks of the downloaded file, not derived constraints. No values were averaged across unlike units, sample dates, or result statuses. The `Oxygen` field contains both concentration and saturation units, so its combined numeric range is intentionally flagged as non-comparable. ### D. Specific unresolved WQP anomaly The file reports a preliminary dissolved `Oxygen` observation of **674 mg/l** at `USGS-05536356` on **2022-08-30 07:40 EST**, activity `nwisin.01.02200882`. The row identifies a routine sample, stable/normal stage, USGS parameter code `82398`, multiple-vertical collection, US WBH-96 weighted-bottle sampler, and USGS-WRD as the conducting organization and laboratory. No detection condition or qualifier is supplied in that row. The value is retained exactly as reported and is not treated as a validated environmental condition. It requires method/source verification before any aggregation or interpretation. ### E. Internet Archive historical anchor: Gary, 1956 The Internet Archive item **Gary’s Golden Jubilee 1906–1956**, published by the Gary Post-Tribune in 1956, is a local souvenir/history publication about Gary, Indiana, Lake County, and the city’s first fifty years. The item metadata identifies OCR generated from scans and provides the identifier `garys-golden-jubilee-19061956-images`. Page-level observations, checked against rendered PDF pages: 1. **PDF p. 7, “Gary Works Planted Seed of Jubilee City”:** the publication states that work started on March 12, 1906; describes the construction of Gary Steel Works on former sand/swamp land; reports diversion of the Grand Calumet flow into a man-made channel of approximately two miles; and states that more than 12 million cubic yards of sand were excavated, with 4 million cubic yards used as backfill. The same page describes a plant area of over 1,400 acres in its aerial caption and reports that annual rated capacity had risen from 2,750,000 ingot tons to nearly 7 million tons, with employment “over 20,000” at the time of publication. The scan’s OCR is visibly noisy; the page image is the controlling check for the bounded wording. 2. **PDF p. 4, historical overview:** the publication reports that during W. P. Gleason’s 27-year tenure as head of the mills, the Gary plant produced more than 56,000,000 tons of steel and describes it as the largest single steel plant in the world. This is a local publication’s retrospective claim, not an independently audited production series. 3. **PDF p. 4, community context:** the publication says Gary may have reached the 170,000 population mark and describes thousands of wage earners and labor-union members. The phrase “may now have reached” is retained as the source’s uncertainty; it is not converted into an exact population value. These observations document the period’s locally narrated industrial scale and infrastructure transformation. They do not prove that the entire fixed sector had the stated production, employment, or population, and they do not establish causation between industrial growth and water conditions. ## Source interpretation - The modern WQP profile provides a reproducible, machine-readable observation set at a named Grand Calumet River station, with explicit dates, units, statuses, qualifiers, and method fields. - The historical Gary publication provides dated institutional and infrastructural context: steel-plant construction, channel modification, rail/harbor integration, industrial land transformation, and locally reported employment/population scale. - The two evidence systems converge only at the level of a **related industrial-water corridor context**. They are not directly comparable as measurements: the historical source is a local retrospective and the WQP source is a point-monitoring profile collected many decades later. - The historical source is geographically related to `N05_E18` but does not cover the full cell. The WQP station is a point in the Grand Calumet watershed and is not an area-weighted cell measure. ## Derived profile (not a quadrant) ### Derivation D1 — WQP profile completeness check - **Input:** all 849 returned water rows in the exact-site CSV. - **Operations:** count rows, distinct dates, characteristic names, result statuses, detection conditions, and numeric ranges within each characteristic/unit grouping. - **Missing-data behavior:** no imputation; no zero substitution; no pooling across unlike units or result statuses. - **Output:** a bounded station-profile summary only. ### Derivation D2 — historical/modern crosswalk | Historical observation | Modern observation | Comparability | Limitation | |---|---|---|---| | Gary Works / Grand Calumet channel modification and industrial footprint, 1906–1956 | Grand Calumet station location and water-quality results, 2021–2024 | Related but not equivalent | Different dates, source boundaries, variables, and observation designs | | Local publication’s reported plant capacity and employment | WQP result counts and analyte values | Not comparable | Production/employment are not water-quality measurements | | Local description of watercourse modification | Station-level stream monitoring | Qualitative/related | Historical hydrologic change requires engineering and map records, not inference from later samples | No constraint score, adaptive-capacity score, or normalized value is calculated in this note. ## Model classification ```yaml provisional_quadrant: unresolved classification_reason: > The run adds a provenance-bearing 1956 historical industrial anchor and a modern point-monitoring profile, but they remain heterogeneous in geography, time, variable type, and measurement design. The WQP oxygen anomaly is not resolved, and the pilot formula and cell-level crosswalk are not locked. formula_version: not_assigned confidence: medium for the named source observations; low for sector-wide inference ``` `provisional_quadrant` is a model output, not an archival fact. This run assigns none. ## Confidence and missingness ### Confidence dimensions - **Source authority:** high for WQP metadata/results as records of what the portal reports; medium for the Gary publication as a local primary-era institutional account and historical source. - **Geographic fit:** medium-low. The station is a point; the Gary publication describes a municipality/industrial corridor; neither equals the fixed cell. - **Temporal fit:** medium for the dated 1956 and 2021–2024 anchors; low for continuous 1950–2000 reconstruction. - **Measurement quality:** high for the transmitted WQP fields, subject to the preliminary status and unresolved 674 mg/l oxygen row; not applicable as a measured series to the historical narrative. - **Extraction quality:** high for CSV counts and field-preserving summaries; medium for OCR because the scan is visually required for page-level claims. - **Independent corroboration:** low to medium. The publication’s industrial-water context is consistent with the already recorded 1965 pollution report and 1968 planning study, but this run does not independently audit those sources again. - **Comparability:** low. The historical and modern evidence are related context, not directly comparable variables. - **Known bias:** the 1956 publication is civic/promotional local history; WQP coverage reflects agency monitoring design, reporting availability, and later sampling priorities. ### Missingness codes - `historical_measure_not_available`: no continuous cell-level 1950–2000 water-quality or industrial-employment series was assembled. - `geographic_match_uncertain`: Gary, the Grand Calumet station, the watershed, and the fixed cell are different footprints. - `period_match_uncertain`: the historical publication is a 1956 retrospective; WQP observations begin in 2021 for this returned profile. - `conflicting_sources`: not established, but the 674 mg/l preliminary oxygen row conflicts with the scale of the other dissolved-oxygen concentration observations and requires source-method review. - `source_exists_not_accessible`: engineering, municipal, utility, industrial, labor, and boundary records remain to be located and verified. - `not_searched`: no new extraction for population, employment, energy, rail/port topology, remediation, or Indigenous-authored/primary records was completed in this run. - `not_applicable`: quadrant and composite score are not applicable until the evidence and formula conditions are met. Portal silence and archival silence are not zeros. Source-reported zero or nondetection values, where present, remain source observations and are not converted into clean-condition or constraint scores. ## Falsification and alternative explanations The working thesis would be weakened if: - later engineering records show that the Grand Calumet channel and industrial pattern were driven by corporate/site-selection decisions without a distinctive sector mechanism; - the apparent historical industrial-water relationship does not survive a principled boundary crosswalk; - neighboring Great Lakes sectors with comparable freshwater and transport access show different trajectories under similar institutions; - the WQP anomaly or broader profile proves to be a data/method artifact rather than a usable observation; - the local publication’s industrial scale claims cannot be corroborated by independent production, census, employment, or engineering records; - documentation density is mistaken for adaptive capacity, especially where labor, Indigenous, or local records are missing. ## Next search path 1. Resolve the 2022-08-30 `Oxygen = 674 mg/l` row through the USGS NWIS source/method record and preserve both the original WQP row and the verification result. 2. Retrieve the same WQP analyte-level profile for additional Great Lakes pilot stations and map each point to the fixed grid with explicit boundary uncertainty. 3. Add machine-readable Census population and labor/employment series for Lake County, Gary, Hammond, and Merrillville where boundary vintages permit; do not treat county/place values as cell values. 4. Locate independent USGS, U.S. Steel, municipal, harbor, or engineering records to corroborate the 1956 publication’s production, employment, channel, and land-transformation claims. 5. Search Archive.org for 1950–2000 rail, port, utility, industrial, and pollution-control records, then extend selected anchors backward to 1800–1950. 6. Add Indigenous-authored or primary Great Lakes records before interpreting the earlier colonial and pre-industrial periods. 7. Keep `provisional_quadrant: unresolved` until raw variables, crosswalks, missingness, alternative explanations, and the pilot formula satisfy the SOP. ## Provenance register ### Modern structured source 1. U.S. EPA, USGS, and National Water Quality Monitoring Council, **Water Quality Portal result CSV**, exact site `USGS-05536356`, retrieved 2026-08-06. The returned file contained 849 water rows, 40 distinct dates, 33 characteristic names, 624 `Preliminary` rows, 225 `Accepted` rows, and 12 `Not Detected` result-condition rows. The station coordinates and metadata were taken from the matching station CSV. - Station: - Results: - Portal: - Web services documentation: ### Historical source 2. Gary Post-Tribune, **Gary’s Golden Jubilee 1906–1956**, 1956, Internet Archive identifier `garys-golden-jubilee-19061956-images`. The scan’s OCR derivative was used to locate passages; PDF pages 4 and 7 were rendered and visually checked. Item metadata identifies the creator, date, Lake County/Gary subjects, OCR, and scan derivatives. - Item: - OCR derivative: - PDF: - Metadata: ### Related project sources not re-extracted in this run 3. Prior N05_E18 notes containing the 1965 pollution report, 1966 Lake County context, 1968 planning study, NOAA water-level series, and GLERL lake-temperature data. These are related sources already documented in the collection; this note does not duplicate their extraction. - `2026-08-06-N05-E18-1950-present-water-industrial-bridge.md` - `2026-08-06-N05-E18-water-quality-and-industrial-context.md` - `2026-08-06-N05-E18-water-quality-monitoring-bridge.md` ## Record limits This note is a bounded N05_E18 research result, not a completed sector audit. It establishes no final quadrant, no causal claim, no cell-wide production or water-quality estimate, and no global coverage. It records the exact next verification path so the Great Lakes pilot can continue without duplicating completed work.