Revenue and Royalty Data
A source-backed study of oil and gas public-data download provenance, with reproducible methods, controls, limitations, and land-work applications.
10 official source routes
6 minimum evidence controls
Checked September 17, 2026
Which provenance fields are needed to reproduce a land-research conclusion when an agency download is updated, corrected, replaced, or published on a different cadence from its source documents?
A reproducible conclusion depends on the exact file vintage and transformation record, not merely the agency name and a current landing-page link. Official portals commonly pair live search interfaces with downloadable files, data dictionaries, maps, and documentation. Some downloads are refreshed on stated schedules; others are static or explicitly subject to correction. A later download may therefore differ from the file used in an earlier review even when the URL is unchanged.
This is a study of public-record research controls, not a title opinion, regulatory interpretation, engineering conclusion, survey, or determination about a particular well or interest. It evaluates what the official entry points say they provide and how a support team can preserve the evidence needed for professional review.
Land work often joins evidence that was not designed as one database. A well may have a current structured row, several scanned forms, a map feature, an order, a case file, and a transfer record. Each source can be accurate within its own purpose while using a different identifier, status vocabulary, event date, or update cycle. Treating all of those observations as one undifferentiated “record” hides the very conflicts a reviewer needs to see.
Remote land support adds value by making the research reproducible. The useful deliverable is not a confident summary without lineage. It is a packet that lets a second person repeat the search, identify the file vintage, reopen the material documents, inspect the transformation rules, and understand which questions remain reserved for land, legal, regulatory, engineering, survey, or accounting professionals.
The review used eight official federal or state publisher entry points. Each page was checked on September 17, 2026. The unit of observation was one publisher, one interface or file, one query or retrieval, and one access time. The review recorded the page title, URL, stated record family, supported identifiers, date labels, update statement, output type, and visible coverage limitation.
No production asset sample was used, and no error rate is claimed. Portal descriptions and exposed controls establish how a route can be researched; they do not prove completeness for every jurisdiction, period, well, organization, or document. Dynamic results can change after access, which is why the method treats the retrieval itself as a dated observation.
The minimum control set was:
| Step | Required field | Quality rule |
|---|---|---|
| 1 | publisher and dataset title | Retain the source value and locator before normalization |
| 2 | landing and final download URLs | Retain the source value and locator before normalization |
| 3 | displayed update note and retrieval time | Retain the source value and locator before normalization |
| 4 | file size and cryptographic hash | Retain the source value and locator before normalization |
| 5 | dictionary or schema version | Retain the source value and locator before normalization |
| 6 | transformation log and output hash | Retain the source value and locator before normalization |
Official portals commonly pair live search interfaces with downloadable files, data dictionaries, maps, and documentation. Some downloads are refreshed on stated schedules; others are static or explicitly subject to correction. A later download may therefore differ from the file used in an earlier review even when the URL is unchanged.
The official pages also show why a route name is not enough metadata. A “well search” may expose current attributes, a document image, a downloadable table, or a link into another application. Those outputs are not interchangeable. A map popup can support discovery of a candidate API number, but the underlying filing is stronger evidence for the text and date on that filing. A downloadable row supports a structured comparison, but its presence does not prove that the image set is complete.
Publisher cadence is likewise descriptive rather than dispositive. Nightly, daily, weekly, and static labels tell the researcher when a publisher expects to refresh a product. They do not establish when an operator submitted an event, when staff accepted it, when an image became visible, or whether a later correction will replace the value. Those are separate fields and should remain separate in the working model.
Match in descending order of strength. First use the issuing jurisdiction and its stable identifier. Next use explicit cross-references inside documents or official linked systems. Then use a combination of well name, operator name, legal location, dates, form types, pool, and map position as corroboration. Names, nearby dates, acreage, or geographic proximity alone remain candidate evidence.
Identifiers must be stored as text. Leading zeroes, punctuation, suffixes, and jurisdictional prefixes can carry meaning or prevent spreadsheet damage. If a normalized identifier is required for joining, retain the original beside it and document the transformation. A failed join is an exception to investigate, not permission to shorten the source values until they match.
Current displays need a current-as-observed label. They should not be projected backward to fill a historical gap. When the record suggests a transfer or amendment, build a chronology from separately sourced events: submitted, received, filed, approved, effective, imaged, published, and accessed. If an intermediate event is unavailable, leave the interval open.
Dates are useful only when paired with the event they describe. Document execution, filing, receipt, approval, effective date, scan date, publication, dataset update, retrieval, and professional review can all be different. A generic “record date” field makes later audit impossible.
Statuses require the same discipline. Preserve the original word, code, capitalization, and source. Put any internal grouping in a separate normalized field with a rule and version. A team may group several agency labels for workload reporting, but that convenience should never overwrite the wording used by the publisher or imply legal equivalence.
Negative results need especially careful language. The supportable statement identifies the interface, record family, query values, filters, access time, and returned condition. It does not claim that a record does not exist. The next route may be another official interface, an archival request, a county source, a paid system, a client file, or escalation to the authorized professional.
At retrieval, record the landing page, final file URL, publisher label, retrieval time, displayed update note, byte size, hash, and dictionary version. Preserve source columns unchanged. Every parse, type conversion, join, filter, and derived field needs a reversible rule and an output hash.
For a daily workflow, triage newly retrieved files, changed source displays, deadline-sensitive items, and exceptions returned by reviewers. Give every open item one owner, one next action, and one review date. “Requested” is a workflow state; it is not evidence received. “Ready for review” means the packet is assembled; it does not mean the substantive question is resolved.
A handoff should begin with the question the reviewer must answer. It should include source-native identifiers, direct locators, material pages, original values, normalized values, date meanings, query history, returned conditions, and a concise conflict table. Convenience copies should be labeled with their source and retrieval time so they cannot be mistaken for a controlled original.
A second researcher should reopen a deliberate sample of material locators and confirm the jurisdiction, identifier, displayed value, status wording, event-date meaning, page reference, and retrieval date. They should reproduce important transformations from the source value and confirm that each candidate match states its evidence basis.
Conflicts deserve their own sample. Both observations remain visible, the variance is described at field level, and any selected working value is marked provisional until the named professional accepts it. If a source has changed, add a new dated observation instead of silently rewriting the earlier row.
File-level controls should include a stable filename, retrieval timestamp, publisher, source URL, byte count, and SHA-256 hash. Derived tables should point to the source files and transformation script or written rule. These controls do not certify that the publisher's data is correct. They prove which material the team used and how it produced the working result.
This review was limited to publicly reachable official entry points on September 17, 2026. It did not test authenticated functions, every possible query combination, paid county systems, agency back-office records, client documents, field conditions, or the legal sufficiency of any filing. Search applications can be unavailable, rate-limited, revised, or reorganized after access.
Published update statements do not measure latency at the individual-record level. A dataset described as daily may contain events with different internal processing dates. An image may appear after a structured row or vice versa. Static and legacy files may remain authoritative for a historical question even when they are not refreshed.
The method therefore supports bounded conclusions: what a named source displayed or made available through a documented route at a particular time. It does not support a universal completeness claim or a decision about title, ownership, regulatory compliance, boundaries, operating authority, physical condition, revenue entitlement, or payment accuracy.
For oil and gas public-data download provenance, defensible research preserves source identity, query path, original identifiers, event-specific dates, file vintage, and inference limits. The strongest land-support packet makes conflicts and unavailable evidence easy to see. It does not manufacture agreement among systems or convert an administrative search into a professional determination.
No. It may support a candidate relationship, especially when jurisdictional identifiers and explicit cross-references agree. The legal or professional significance remains with the authorized reviewer.
State the publisher, interface, record family, exact query and filters, access date, and returned condition. Then identify the next approved route. Do not turn one negative query into a claim that no record exists.
Yes. Trained support staff can retrieve authorized sources, preserve files, transcribe source values, apply documented transformations, maintain exception queues, and prepare review packets. Reserved legal, land, regulatory, survey, engineering, and accounting decisions must stay with qualified professionals.
If your team needs help building a repeatable oil and gas public-data download provenance workflow, Book a free consultation.
A decision-grade, source-backed study of ONRR Form 2014 adjustment and reporting-period lineage, with reproducible controls, limitations, and land-work applications.
Verified 2026-09-23
A decision-grade, source-backed study of ONRR lease and agreement number crosswalk lineage, with reproducible controls, limitations, and land-work applications.
Verified 2026-09-23
A source-backed study of ONRR royalty adjustment and dataset-version research, with reproducible controls, limitations, and land-work applications.
Verified 2026-09-18
