The change compares one month with the same month a year earlier, or one year with the preceding year. It is descriptive and does not predict the next period.
United States · national series
U.S. marketed natural gas production
Million cubic feet
Latest reported period ·
41,383,894.00
Change from the same period a year earlier
+0.8%
· 41,037,738.00
1960–2024 · 65 Retained periodsRecent values and their coverage
U.S. marketed natural gas production · Recent values and their coverage
Source: Statistics Canada, Crude oil and equivalent, monthly supply and disposition, reference dates 1985-01 through 2016-02. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Source acquired:
Canada · national series
Canada crude oil production
Cubic metres
Latest reported period ·
24,876,835.00
Change from the same period a year earlier
+2.3%
· 24,325,022.00
2016-01–2026-07 · 127 Retained periodsRecent values and their coverage
Canada crude oil production · Recent values and their coverage
Source: Statistics Canada, Supply and disposition of crude oil and equivalent, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Source acquired:
Canada · national series
Canada marketable natural gas production
Cubic metres · Native scale multiplier 103
Latest reported period ·
17,568,809.80
Change from the same period a year earlier
+4.2%
· 16,859,814.40
2016-01–2026-07 · 127 Retained periodsRecent values and their coverage
Canada marketable natural gas production · Recent values and their coverage
Source: Statistics Canada, Natural gas supply and disposition, monthly, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Source acquired:
BSEE · reported offshore completions
BSEE reported oil volumes
bbl
Latest reported period ·
57,419,970.00
Change from the same period a year earlier
-1.8%
· 58,469,587.00
1996-01–2026-06 · 366 Retained periodsRecent values and their coverage
BSEE reported oil volumes · Recent values and their coverage
Reported period
Source value
Reported cells used
Missing cells
Quality flags
57,419,970.00
4,365
0
source_reported_zero
59,380,671.00
4,433
0
source_reported_zero
63,955,183.00
4,430
0
source_reported_zero
62,994,915.00
4,637
0
source_reported_zero
56,956,960.00
4,663
0
source_reported_zero
64,971,130.00
4,673
0
source_reported_zero
65,078,743.00
4,705
0
source_reported_zero
62,095,590.00
4,726
0
source_reported_zero
66,501,289.00
4,735
0
source_reported_zero
61,931,894.00
4,744
0
source_reported_zero
62,996,525.00
4,758
0
source_reported_zero
59,986,591.00
4,798
0
source_reported_zero
Sum of source completion cells
Exact source selection: {"grain":"source completion/month including allocation identity","historicalAvailableAt":null,"latestYearIsPartial":true,"officialReconciledProductionTotal":false,"physicalWellCount":false,"retainedExactZeroDuplicateRows":1}
Source: U.S. Bureau of Safety and Environmental Enforcement (BSEE), OGOR-A Production 2025, downloaded directly from the BSEE Data Center.
Source acquired:
What these comparisons mean
Every retained point in this source selection is included. Missing values and gaps break the line; values are not filled or clipped. Each chart has its own axis.
National, state and reported-completion series have different scopes. Commodity definitions differ between production and consumption, so their difference is not a net-import measure. Historical and current Statistics Canada oil tables remain separate. Producing-well counts are monthly source aggregates and are not added across months.
BSEE values sum reported completion-month rows, including separate completion records. They are not a reconciled count of physical wells or an official cross-source production total.
Document search covers all news and source documents; it does not filter measurements.
Browse measurements
Measurements
475,830 records
Runtime samples are not the complete training dataset. Depth filters use source units; select one dataset first. Date filters include observations whose reported period overlaps the selected range.
Source: U.S. Bureau of Safety and Environmental Enforcement (BSEE), OGOR-A annual production archives, downloaded directly from the BSEE Data Center.
Current revised annual snapshots, not historic available-at vintages. Source row grain is completion/month; API sidetracks and completions are not independently verified physical wells. No aggregation, imputation, unit conversion or cross-source matching. 2025 bytes are verified but normalized rows are excluded here to avoid duplicating the separate unchanged bsee-ogora-well-production-2025 dataset. 2026 is incomplete and its last supplied month is reported in coverage. Raw complete records remain in source ZIPs; public measurement payloads preserve exact row pointers and production context.
Source: U.S. Bureau of Safety and Environmental Enforcement (BSEE), OGOR-A Production 2025, downloaded directly from the BSEE Data Center.
The live archive is a current annual snapshot, not a preserved historical vintage. Blank volumes are unknown/missing and remain NULL; source zeros remain numeric zero. No physical-well deduplication, aggregation, or joins to separate lease/borehole files are performed. BSEE/BOEM disclaim data accuracy/completeness; public copying language does not establish separate commercial reuse or ML terms for the underlying reporting records.
this uses Environment Agency flood and river level data from the real-time data API (Beta)
This is the 2,784 source readings returned for one Environment Agency groundwater station; the requested September 1 start was clipped by source retention to September 6. It is not a source-wide coverage claim. Acquisition time is not historical availability. Station IDs do not identify physical wells. Values are groundwater level relative to mAOD, not pumping or flood labels. No forecast or model performance is claimed.
Irregular instantaneous dipped measurements in native mAOD, not depth below land surface or daily means. No interpolation or zero filling. Bounded station subset; source GUIDs are not physical-well reconciliation. Source timezone, historical availability and parsed coordinate CRS remain unknown. Original bounded first-1000 station list, sorted source GUIDs, qualified dipped histories spanning at least one year. Original selection probes were not retained; proof covers retained import responses only.
Source: U.S. Energy Information Administration (EIA), State Energy Data System (SEDS), release 2026-06-26.
Complete allowlisted annual series from the official SEDS bulk source. Geography includes U.S. national and state/DC records plus only the identified federal offshore Gulf of America (X3) and Pacific (X5) production regions. Offshore regions are separate geography entities, not state rows; the U.S. total can include federal offshore production and need not equal the sum of states. State estimates are energy-system statistics, not well-level production. Native annual quantity, period, measure, and unit are retained. Production/consumption estimates may be statistical estimates described in SEDS technical notes. Historical availableAt is unknown; acquisition time is not historical vintage availability.
Credit Equinor and former Volve licence partners (ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS), link the terms, and include copyright notice if applicable.
Equinor-published Volve BLOCKING example subset: processed/interpreted curves, not raw sensor logs or the complete Volve archive. Source identifiers include sidetracks and are not automatically matched to registry wells. Missing units, depth datum and observation times remain unknown. Discrete codes are categories, not continuous measurements. Runtime retains first/last and every 180th depth row; no interpolation. Full source cells are retained in a separate external normalized snapshot.
McLennan, John (2021), Utah FORGE: Well 78B-32 Daily Drilling Reports and Logs, Energy and Geoscience Institute at the University of Utah, Geothermal Data Repository, DOI 10.15121/1814488.
Selected LAS files only; the 22-file, 11.89 GB parent submission was not bulk-imported. Source well, vendor, depth and curve units, null sentinel, log date and repeated-run provenance are preserved. Schlumberger headers have no explicit proprietary exclusion in these selected files; GDR CC BY 4.0 and attribution apply. No depth-datum or coordinate conversion, averaging or interpolation.
McLennan, John (2021), Utah FORGE: Well 78B-32 Daily Drilling Reports and Logs, Energy and Geoscience Institute at the University of Utah, Geothermal Data Repository, DOI 10.15121/1814488.
Selected LAS files only; the 22-file, 11.89 GB parent submission was not bulk-imported. Source well, vendor, depth and curve units, null sentinel, log date and repeated-run provenance are preserved. Schlumberger headers have no explicit proprietary exclusion in these selected files; GDR CC BY 4.0 and attribution apply. No depth-datum or coordinate conversion, averaging or interpolation.
McLennan, John (2025), Utah FORGE: Shut-in Wellhead Pressure Data for Wells 16A(78)-32 and 16B(78)-32, Energy and Geoscience Institute at the University of Utah, Geothermal Data Repository, https://gdr.openei.org/submissions/1764.
Discontinuous wellhead-pressure windows, not downhole/reservoir pressure. Units PSI; source timezone is unreported. Literal -999.25 sentinels are NULL with quality flags and original source values preserved. Gaps and original timestamps retained; no interpolation or reconstructed history.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
KGS source is county and statewide monthly aggregate production and producing-well counts, not per-well production or telemetry. Do not add statewide totals to the county dataset. The current source states data are through May 2026; later monthly cells are retained as source nulls and are not treated as observed. Values are preserved as reported without unit conversion, interpolation or gap filling.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
A bounded sample of eight TestDataSave rows for three source KIDs, not a statewide or complete KGS DST collection. KGS KID/API identifiers are source identifiers, not independently resolved physical-well identities. Bottom-hole temperature comes from CSV rows with explicit DEG F. Pressure summary cells come only from detail-page blocks uniquely matched by KID, test number, date, and time; the page labels PSI and the matched CSV declares PSIg. CSV depth values have no unit field and remain raw metadata only. Source date and time fields are kept separately; observation date has day precision and source time has no documented timezone. Linked vendor PDF is archive-only and excluded from normalized content; pressure curves are omitted.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
Well-level DIGITAL/SCANNED report availability, not actual DST test measurements or downloaded report contents. Event dates, depths and coordinates in the original index have not been promoted to test observations or verified spatial measurements. No current well status or physical-well cross-source matching is inferred. Both availability kinds for the same KID remain distinct.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
KGS source is county and statewide monthly aggregate production and producing-well counts, not per-well production or telemetry. Do not add statewide totals to the county dataset. The current source states data are through May 2026; later monthly cells are retained as source nulls and are not treated as observed. Values are preserved as reported without unit conversion, interpolation or gap filling.
Source: NLOG / TNO Geological Survey of the Netherlands.
Selected pore-fluid pressures underlying overpressure mapping, from activities through 2012. Heterogeneous tests and interpreted values with source reliability/remarks; not current telemetry. Only Pressures_Total imported: stratigraphic sheets overlap and are not double-counted. Depth is TVDSS in m; coordinates remain source UTM31 with datum unverified and are not mapped as WGS84.
NLOG / TNO Geological Survey of the Netherlands; cite NLOG as the source and state modifications.
Official NLOG Datacenter well production/injection output, source-supplied wide monthly cells grouped by source well name, field and operator, 2003–2026. The API response has no stable well identifier; the composite source key used for grouping is not a reconciled physical-well identity. A returned zero remains zero; source null remains null; no absent month is filled. Values are per API product and operation, not summed across products. Monthly periods have YYYY-MM precision; future month columns have null observation dates with their source period retained separately. Actual UTC acquisition time is separate from the server Date header and historical availability. Raw JSON does not declare units. UI unit selectors and gas scaling do not authenticate raw response units; normalized units remain unknown pending a matched official data export. Recent months may be provisional or not yet supplied; the provider says monthly data are published after a reporting lag. This is not sensor telemetry or a forecast label.
Source: NOAA National Centers for Environmental Information (NCEI), Global Historical Climatology Network Daily (GHCN-Daily).
Bounded deterministic subset of at most 20 contiguous-US stations selected nearest a fixed set of broad regional targets among stations whose inventory spans 2016–2025 for PRCP/TMAX/TMIN; not source-wide coverage. Rows are daily station observations and climate covariates, not well, groundwater, precipitation-at-well, or recharge measurements; no well matching or spatial equivalence is asserted. Native integer values and NOAA element units are retained without conversion: PRCP is tenths of mm; TMAX/TMIN are tenths of degrees Celsius. Missing sentinel -9999 is null with provenance; PRCP measurement flag T is null rather than treated as zero. Quality, measurement, and source flags are retained verbatim as flags and raw records; no quality-flag filtering is applied. Station metadata provides latitude, longitude, and elevation in metres (-999.9 means missing); parsed coordinates are not reprojected and CRS datum is not asserted. Daily source date is retained without asserting a timezone or observation boundary. Historical availableAt is unknown. Retrieval time records this acquisition, not when each historical observation became available; source archive may change. GHCN-D combines multiple source networks and QA-reviewed records; daily dates may follow local/source reporting conventions, not a globally synchronized day. Source documentation disclaims accuracy/reliability/completeness warranties; use requires independent suitability review.
Contains information provided by the North Sea Transition Authority and/or other third parties.
Historical disclosed production, not current telemetry. All 182 index entries collected as 181 unique ZIPs (one shared URL retained once); runtime imports sampled original cells from 150 CSV files. The 31 Excel workbooks are imported separately in nsta-ukcs-daily-workbooks-sample with reviewed per-well scope and units. No daily/monthly aggregation, gap filling, unit conversion or physical-well matching. CSV headers and source units retained; Buchan units are explicit in column names.
Contains information provided by the North Sea Transition Authority and/or other third parties.
Historical disclosed daily records from 31 per-well post-COP workbooks. Not current telemetry. Only rows and columns explicitly tied to the workbook well are considered. Platform/field totals, unallocated platform streams and inferred values are excluded. Shell/Heron imports only allocated per-well fluid columns and online hours. Banff platform choke and injected-gas density columns are excluded because their well-level scope or units are unclear. Source dates, raw values and units are preserved; no unit conversion, interpolation or deduplication between source files is performed. Some workbooks publish the same quantity in separate unit columns (for example Sm3 and MMSCF); those columns are alternate representations and must not be summed. Identifiers name source wellbores and are not asserted to represent unique physical wells.
Source: Government of Alberta, Alberta Public Data / Petrinex, Alberta volumetric reporting, production month 2025-01. Crown copyright acknowledged. This is a reproduction/adaptation and is not an official version or affiliated with or endorsed by the Government of Alberta, AER or Petrinex.
ProductionMonth is monthly precision; source periods outside the archive label are retained and flagged.,Facility, activity, product and well identifiers retain their source scope; none is asserted to be a reconciled physical-well identity.,Unknown product-specific volume units remain null with an explicit flag.,No model training, accuracy, or forecast claim is made.
Petrobras / 3W Project; cite dataset v2.0.0 and the pinned source instance.
Five deliberately small real instances from five anonymized source-well groups, not the complete or balanced 3W dataset. Simulated/drawn instances are excluded. Naive source clocks have no verified timezone; public observation dates have day precision with exact source clocks in the payload. Source identity is not a reconciled physical well. Label/state roles and codebooks are explicit; exclude target labels from predictors. Provider labels are not predictions, failure onset or universally transferable ground truth. Historical availability, calibration and production model accuracy are not established.
Contains data under the Norwegian licence for Open Government data (NLOD) distributed by the Norwegian Offshore Directorate.
Source test/engineering summaries, not daily production or current well status. Source QC and update dates are not measurement times; observedAt stays null. Source zeros are retained and flagged as unqualified, not changed to missing. Diameter source strings are retained; decimal and fractional inches are parsed exactly, while uninterpreted labels stay null with a flag. No geographic or physical-well matches are inferred.
Contains data under the Norwegian licence for Open Government data (NLOD) distributed by the Norwegian Offshore Directorate.
Source test/engineering summaries, not daily production or current well status. Source QC and update dates are not measurement times; observedAt stays null. Source zeros are retained and flagged as unqualified, not changed to missing. Diameter source strings are retained; decimal and fractional inches are parsed exactly, while uninterpreted labels stay null with a flag. No geographic or physical-well matches are inferred.
Source: Statistics Canada, Supply and disposition of crude oil and equivalent, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Source: Statistics Canada, Crude oil and equivalent, monthly supply and disposition, reference dates 1985-01 through 2016-02. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Source: Statistics Canada, Natural gas supply and disposition, monthly, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
U.S. Geological Survey, Water Data APIs. Cite the monitoring-location and daily-series records; identify normalization changes.
This is a bounded source snapshot, not all USGS groundwater observations. Daily values for parameter 72019 are depth below land surface; only daily mean statistic 00003 series are selected. Native units remain attached. Source last_modified is retained as record refresh metadata, not historical availability. Today’s API response cannot establish when a past observation was first publicly available. Site IDs identify monitoring locations, not reconciled physical wells. Only agency_code=USGS records are included.
U.S. Geological Survey, NWIS records retrieved through the USGS/EPA Water Quality Portal; indicate normalization changes.
Only source stations matching the collected chemistry records are normalized; the complete downloaded Texas station CSV remains external. Well/hole depths and site elevation are distinct; source units, horizontal datum, vertical datum, aquifer, well type and source construction-date text are preserved. A construction date is not a depth measurement date or historical availability; unknown observation dates stay null. NAD83 is not asserted to be WGS84, and no coordinates are converted.
U.S. Geological Survey, NWIS records retrieved through the USGS/EPA Water Quality Portal; indicate normalization changes.
Historical WQX 2.2 NWIS response for Texas groundwater chemistry. This profile excludes USGS additions after March 11, 2024; requested date bounds do not establish actual coverage. Concentration fractions, analytical methods, detection conditions, quantitation limits, qualifiers and QC samples must be reviewed before training. Non-detects are null, not zero or the quantitation limit. Source agency/site/activity identifiers do not reconcile physical wells. Acquisition is not historical availability. Source clock time and timezone code are retained without inventing a UTC instant.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Workbook mirror supplied by Jamiu Olamilekan Badmus via Kaggle; not fetched from Equinor.
Historical Volve workbook from a pinned public third-party mirror, not directly authenticated against Equinor. Source zero/null, flow kind and well type are retained, including inconsistent row labels. Daily columns do not declare most units: these remain unknown; monthly sheet explicitly gives hours and Sm3. Monthly dates retain month precision. Daily and monthly sheets overlap and must not be summed together. No forecast, current status or point-in-time availability is inferred.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Workbook mirror supplied by Jamiu Olamilekan Badmus via Kaggle; not fetched from Equinor.
Historical Volve workbook from a pinned public third-party mirror, not directly authenticated against Equinor. Source zero/null, flow kind and well type are retained, including inconsistent row labels. Daily columns do not declare most units: these remain unknown; monthly sheet explicitly gives hours and Sm3. Monthly dates retain month precision. Daily and monthly sheets overlap and must not be summed together. No forecast, current status or point-in-time availability is inferred.
Bormann, P.; Aursand, P.; Dilib, F.; Manral, S.; Dischington, P. (2020). FORCE 2020 well log and lithofacies dataset, DOI 10.5281/zenodo.4351156. Source Norwegian government logs under NLOD 2.0. Normalized and sampled by Wellficent.
Runtime sample is for browsing. Full source-row normalized training data must use external importer output.
Data gaps and limitations
Unknown units remain unknown. Undated depth logs are not automatically defective records.
Current revised annual snapshots, not historic available-at vintages. Source row grain is completion/month; API sidetracks and completions are not independently verified physical wells. No aggregation, imputation, unit conversion or cross-source matching. 2025 bytes are verified but normalized rows are excluded here to avoid duplicating the separate unchanged bsee-ogora-well-production-2025 dataset. 2026 is incomplete and its last supplied month is reported in coverage. Raw complete records remain in source ZIPs; public measurement payloads preserve exact row pointers and production context.
First/last and every Nth original source row in each annual archive, with all three native quantities including nulls. No aggregation or interpolation. This is a bounded source-row subset, not a representative physical-well sample.
The live archive is a current annual snapshot, not a preserved historical vintage. Blank volumes are unknown/missing and remain NULL; source zeros remain numeric zero. No physical-well deduplication, aggregation, or joins to separate lease/borehole files are performed. BSEE/BOEM disclaim data accuracy/completeness; public copying language does not establish separate commercial reuse or ML terms for the underlying reporting records.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Environment Agency · Kingston Russell groundwater level readings · 2,784
This is the 2,784 source readings returned for one Environment Agency groundwater station; the requested September 1 start was clipped by source retention to September 6. It is not a source-wide coverage claim. Acquisition time is not historical availability. Station IDs do not identify physical wells. Values are groundwater level relative to mAOD, not pumping or flood labels. No forecast or model performance is claimed.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
EA historical qualified dipped groundwater levels · 2,028
Irregular instantaneous dipped measurements in native mAOD, not depth below land surface or daily means. No interpolation or zero filling. Bounded station subset; source GUIDs are not physical-well reconciliation. Source timezone, historical availability and parsed coordinate CRS remain unknown. Original bounded first-1000 station list, sorted source GUIDs, qualified dipped histories spanning at least one year. Original selection probes were not retained; proof covers retained import responses only.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
EIA SEDS · annual U.S. state and federal offshore oil and gas production and consumption · 3,587
Complete allowlisted annual series from the official SEDS bulk source. Geography includes U.S. national and state/DC records plus only the identified federal offshore Gulf of America (X3) and Pacific (X5) production regions. Offshore regions are separate geography entities, not state rows; the U.S. total can include federal offshore production and need not equal the sum of states. State estimates are energy-system statistics, not well-level production. Native annual quantity, period, measure, and unit are retained. Production/consumption estimates may be statistical estimates described in SEDS technical notes. Historical availableAt is unknown; acquisition time is not historical vintage availability.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Equinor-published Volve BLOCKING example subset: processed/interpreted curves, not raw sensor logs or the complete Volve archive. Source identifiers include sidetracks and are not automatically matched to registry wells. Missing units, depth datum and observation times remain unknown. Discrete codes are categories, not continuous measurements. Runtime retains first/last and every 180th depth row; no interpolation. Full source cells are retained in a separate external normalized snapshot.
First/last depth row and every 180th row per BLOCKING source record; categorical codes and nulls retained.
Selected LAS files only; the 22-file, 11.89 GB parent submission was not bulk-imported. Source well, vendor, depth and curve units, null sentinel, log date and repeated-run provenance are preserved. Schlumberger headers have no explicit proprietary exclusion in these selected files; GDR CC BY 4.0 and attribution apply. No depth-datum or coordinate conversion, averaging or interpolation.
source depth rows every 180 samples, plus first/last and depth-gap transition rows
FORGE Well 78B-32 Porosity and Resistivity Logs.zip · 3,400
Selected LAS files only; the 22-file, 11.89 GB parent submission was not bulk-imported. Source well, vendor, depth and curve units, null sentinel, log date and repeated-run provenance are preserved. Schlumberger headers have no explicit proprietary exclusion in these selected files; GDR CC BY 4.0 and attribution apply. No depth-datum or coordinate conversion, averaging or interpolation.
source depth rows every 180 samples, plus first/last and depth-gap transition rows
Utah FORGE shut-in wellhead pressure data for wells 16A(78)-32 and 16B(78)-32 · 47,034
Discontinuous wellhead-pressure windows, not downhole/reservoir pressure. Units PSI; source timezone is unreported. Literal -999.25 sentinels are NULL with quality flags and original source values preserved. Gaps and original timestamps retained; no interpolation or reconstructed history.
every 30th source timestamp row plus first, last, and gap-transition rows
Kansas · monthly county oil and gas production, 1970–2026 · 4,868
KGS source is county and statewide monthly aggregate production and producing-well counts, not per-well production or telemetry. Do not add statewide totals to the county dataset. The current source states data are through May 2026; later monthly cells are retained as source nulls and are not treated as observed. Values are preserved as reported without unit conversion, interpolation or gap filling.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Kansas · selected digital drill stem test measurements · 72
A bounded sample of eight TestDataSave rows for three source KIDs, not a statewide or complete KGS DST collection. KGS KID/API identifiers are source identifiers, not independently resolved physical-well identities. Bottom-hole temperature comes from CSV rows with explicit DEG F. Pressure summary cells come only from detail-page blocks uniquely matched by KID, test number, date, and time; the page labels PSI and the matched CSV declares PSIg. CSV depth values have no unit field and remain raw metadata only. Source date and time fields are kept separately; observation date has day precision and source time has no documented timezone. Linked vendor PDF is archive-only and excluded from normalized content; pressure curves are omitted.
All eight source CSV test rows, retaining source-null temperature and pressure channels. Detail summary cells are imported only from blocks uniquely matched by KID, test number, date and time.
Kansas · drill stem test availability index · 0
Reported values: 0 · Missing values: 0 · Flagged values: Not reported / not verified
Channel preview is limited to 50 entries; use filters to access all records.
Data coverage
Entity
Source field
Source file
Unit
Source depth · Unit
records
Well-level DIGITAL/SCANNED report availability, not actual DST test measurements or downloaded report contents. Event dates, depths and coordinates in the original index have not been promoted to test observations or verified spatial measurements. No current well status or physical-well cross-source matching is inferred. Both availability kinds for the same KID remain distinct.
All source KID/availability rows; only byte-equivalent repeated records are deduplicated, with original row numbers retained. Source well pages remain external.
Kansas · monthly state oil and gas production, 1970–2026 · 2,736
KGS source is county and statewide monthly aggregate production and producing-well counts, not per-well production or telemetry. Do not add statewide totals to the county dataset. The current source states data are through May 2026; later monthly cells are retained as source nulls and are not treated as observed. Values are preserved as reported without unit conversion, interpolation or gap filling.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Selected pore-fluid pressures underlying overpressure mapping, from activities through 2012. Heterogeneous tests and interpreted values with source reliability/remarks; not current telemetry. Only Pressures_Total imported: stratigraphic sheets overlap and are not double-counted. Depth is TVDSS in m; coordinates remain source UTM31 with datum unverified and are not mapped as WGS84.
All nonblank Pressures_Total rows; pressure only, with depth, raw coordinates, fluid/test type and source reliability.
NLOG · monthly well production and injection, Netherlands · 13,783
Official NLOG Datacenter well production/injection output, source-supplied wide monthly cells grouped by source well name, field and operator, 2003–2026. The API response has no stable well identifier; the composite source key used for grouping is not a reconciled physical-well identity. A returned zero remains zero; source null remains null; no absent month is filled. Values are per API product and operation, not summed across products. Monthly periods have YYYY-MM precision; future month columns have null observation dates with their source period retained separately. Actual UTC acquisition time is separate from the server Date header and historical availability. Raw JSON does not declare units. UI unit selectors and gas scaling do not authenticate raw response units; normalized units remain unknown pending a matched official data export. Recent months may be provisional or not yet supplied; the provider says monthly data are published after a reporting lag. This is not sensor telemetry or a forecast label.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
NOAA GHCN-Daily US station daily climate covariates, bounded subset 2016–2025 · 2,986
Bounded deterministic subset of at most 20 contiguous-US stations selected nearest a fixed set of broad regional targets among stations whose inventory spans 2016–2025 for PRCP/TMAX/TMIN; not source-wide coverage. Rows are daily station observations and climate covariates, not well, groundwater, precipitation-at-well, or recharge measurements; no well matching or spatial equivalence is asserted. Native integer values and NOAA element units are retained without conversion: PRCP is tenths of mm; TMAX/TMIN are tenths of degrees Celsius. Missing sentinel -9999 is null with provenance; PRCP measurement flag T is null rather than treated as zero. Quality, measurement, and source flags are retained verbatim as flags and raw records; no quality-flag filtering is applied. Station metadata provides latitude, longitude, and elevation in metres (-999.9 means missing); parsed coordinates are not reprojected and CRS datum is not asserted. Daily source date is retained without asserting a timezone or observation boundary. Historical availableAt is unknown. Retrieval time records this acquisition, not when each historical observation became available; source archive may change. GHCN-D combines multiple source networks and QA-reviewed records; daily dates may follow local/source reporting conventions, not a globally synchronized day. Source documentation disclaims accuracy/reliability/completeness warranties; use requires independent suitability review.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
NSTA · historical UKCS daily well production samples · 65,775
Historical disclosed production, not current telemetry. All 182 index entries collected as 181 unique ZIPs (one shared URL retained once); runtime imports sampled original cells from 150 CSV files. The 31 Excel workbooks are imported separately in nsta-ukcs-daily-workbooks-sample with reviewed per-well scope and units. No daily/monthly aggregation, gap filling, unit conversion or physical-well matching. CSV headers and source units retained; Buchan units are explicit in column names.
First/last and every 180th source row in each reviewed CSV, all measurement columns including nulls; full originals retained externally.
NSTA · post-cessation daily well production workbooks · 14,141
Historical disclosed daily records from 31 per-well post-COP workbooks. Not current telemetry. Only rows and columns explicitly tied to the workbook well are considered. Platform/field totals, unallocated platform streams and inferred values are excluded. Shell/Heron imports only allocated per-well fluid columns and online hours. Banff platform choke and injected-gas density columns are excluded because their well-level scope or units are unclear. Source dates, raw values and units are preserved; no unit conversion, interpolation or deduplication between source files is performed. Some workbooks publish the same quantity in separate unit columns (for example Sm3 and MMSCF); those columns are alternate representations and must not be summed. Identifiers name source wellbores and are not asserted to represent unique physical wells.
First, last, and every 180th dated source row per workbook; all reviewed measurement columns for retained rows. Full originals remain external.
Alberta · January 2025 volumetric reporting archive · 1,828
ProductionMonth is monthly precision; source periods outside the archive label are retained and flagged.,Facility, activity, product and well identifiers retain their source scope; none is asserted to be a reconciled physical-well identity.,Unknown product-specific volume units remain null with an explicit flag.,No model training, accuracy, or forecast claim is made.
Select original CSV row 1, every 400th data row, and the final data row; retain every numeric channel from each selected row. This is a compact sample of source reporting rows, not all wells or well histories.
Five deliberately small real instances from five anonymized source-well groups, not the complete or balanced 3W dataset. Simulated/drawn instances are excluded. Naive source clocks have no verified timezone; public observation dates have day precision with exact source clocks in the payload. Source identity is not a reconciled physical well. Label/state roles and codebooks are explicit; exclude target labels from predictors. Provider labels are not predictions, failure onset or universally transferable ground truth. Historical availability, calibration and production model accuracy are not established.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Source test/engineering summaries, not daily production or current well status. Source QC and update dates are not measurement times; observedAt stays null. Source zeros are retained and flagged as unqualified, not changed to missing. Diameter source strings are retained; decimal and fractional inches are parsed exactly, while uninterpreted labels stay null with a flag. No geographic or physical-well matches are inferred.
All returned structured source rows and reviewed numeric fields; object-ID coverage verified across paginated requests.
Source test/engineering summaries, not daily production or current well status. Source QC and update dates are not measurement times; observedAt stays null. Source zeros are retained and flagged as unqualified, not changed to missing. Diameter source strings are retained; decimal and fractional inches are parsed exactly, while uninterpreted labels stay null with a flag. No geographic or physical-well matches are inferred.
All returned structured source rows and reviewed numeric fields; object-ID coverage verified across paginated requests.
Statistics Canada · crude oil current monthly supply and disposition · 4,960
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Statistics Canada · crude oil and equivalent monthly supply and disposition · 4,574
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Statistics Canada · natural gas monthly supply and disposition · 5,908
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
USGS daily groundwater depth to water level (parameter 72019) · 3,021
This is a bounded source snapshot, not all USGS groundwater observations. Daily values for parameter 72019 are depth below land surface; only daily mean statistic 00003 series are selected. Native units remain attached. Source last_modified is retained as record refresh metadata, not historical availability. Today’s API response cannot establish when a past observation was first publicly available. Site IDs identify monitoring locations, not reconciled physical wells. Only agency_code=USGS records are included.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
USGS · construction and aquifer context for Texas chemistry wells · 4,248
Only source stations matching the collected chemistry records are normalized; the complete downloaded Texas station CSV remains external. Well/hole depths and site elevation are distinct; source units, horizontal datum, vertical datum, aquifer, well type and source construction-date text are preserved. A construction date is not a depth measurement date or historical availability; unknown observation dates stay null. NAD83 is not asserted to be WGS84, and no coordinates are converted.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Historical WQX 2.2 NWIS response for Texas groundwater chemistry. This profile excludes USGS additions after March 11, 2024; requested date bounds do not establish actual coverage. Concentration fractions, analytical methods, detection conditions, quantitation limits, qualifiers and QC samples must be reviewed before training. Non-detects are null, not zero or the quantitation limit. Source agency/site/activity identifiers do not reconcile physical wells. Acquisition is not historical availability. Source clock time and timezone code are retained without inventing a UTC instant.
Actual source first/last and every Nth row per entity/channel/unit/pass; no interpolation or replacement of nulls.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
First/last and every 180th source depth row, all curves, with nulls. Full JSON arrays retained externally for research; no complete normalized SQLite claim.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
First/last and every 180th source depth row, all curves, with nulls. Full JSON arrays retained externally for research; no complete normalized SQLite claim.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
First/last and every 180th source depth row, all curves, with nulls. Full JSON arrays retained externally for research; no complete normalized SQLite claim.
Volve · daily well production (public workbook mirror) · 82,410
Historical Volve workbook from a pinned public third-party mirror, not directly authenticated against Equinor. Source zero/null, flow kind and well type are retained, including inconsistent row labels. Daily columns do not declare most units: these remain unknown; monthly sheet explicitly gives hours and Sm3. Monthly dates retain month precision. Daily and monthly sheets overlap and must not be summed together. No forecast, current status or point-in-time availability is inferred.
All source daily volumes and on-stream hours; other channels retain first/last and every 30th source row per well. Full original workbook retained externally.
Volve · monthly well production (public workbook mirror) · 3,156
Historical Volve workbook from a pinned public third-party mirror, not directly authenticated against Equinor. Source zero/null, flow kind and well type are retained, including inconsistent row labels. Daily columns do not declare most units: these remain unknown; monthly sheet explicitly gives hours and Sm3. Monthly dates retain month precision. Daily and monthly sheets overlap and must not be summed together. No forecast, current status or point-in-time availability is inferred.
All source monthly cells; no temporal interpolation or missing-month filling.
No observation/update date, status, depth or operator is supplied in the imported layer. Coordinate service transforms from EPSG:32636 to EPSG:4326; six decimals are resolution, not accuracy. Personal owner/name fields are excluded.
Source licence provides broad copying, modification and distribution rights but excludes data the licensor cannot license, including third-party rights; preserve source role, purpose, unit and reference metadata. This is a historical/source snapshot, not a live status feed.
Netherlands Organisation for Applied Scientific Research (TNO) / NLOG
EPSG:23031 to EPSG:4326 transform is provider supplied. Original operator, current owner and drilling contractor remain distinct roles. Source public-release date is not treated as a status update.
An authorization is not proof that a well was drilled. Operator abbreviations remain codes. Preserve OGC:CRS84 coordinate order and confidential exclusions.
California Department of Conservation, Geologic Energy Management Division (CalGEM)
Source sentinel coordinates are excluded; status/type/operator codes remain codes. EPSG:3310 transformed to WGS84 without ballpark conversion; this is not a survey-accuracy claim.
Top-hole points only, not well paths or bottom-hole locations. Item-level licence is unspecified; preserve exact NSTA attribution and per-source rights metadata.
Bounded convenience sample, not a national inventory or representative sample. Values are ft below land surface; qualifiers/provisional values and gaps are preserved. USGS public-domain status applies only to USGS-authored records.
U.S. Energy Information Administration; underlying daily source Refinitiv (LSEG)
EIA annual values are unweighted averages derived from daily closing prices supplied by Refinitiv. Daily series are not fetched. EIA reuse terms exclude third-party licensed content.
imported_and_pipeline_archive_registered · CC BY 4.0
Global Heat Flow Data Assessment Group et al. (2024), Global Heat Flow Database Release 2024, v2026.03, GFZ Data Services, DOI 10.5880/fidgeo.2024.014.
Energy and Geoscience Institute, University of Utah; Geothermal Data Repository (DOE GDR)
Discontinuous wellhead-pressure windows, not downhole/reservoir pressure. Units PSI; source timezone is unreported. Literal -999.25 sentinels are NULL with quality flags and original source values preserved. Gaps and original timestamps retained; no interpolation or reconstructed history.
imported_and_pipeline_archive_registered · CC BY 4.0 as displayed on the GDR submission page
McLennan, John (2025), Utah FORGE: Shut-in Wellhead Pressure Data for Wells 16A(78)-32 and 16B(78)-32, Energy and Geoscience Institute at the University of Utah, Geothermal Data Repository, https://gdr.openei.org/submissions/1764.
imported_and_pipeline_archive_registered · CC BY 4.0 as displayed on the GDR submission page
McLennan, John (2021), Utah FORGE: Well 78B-32 Daily Drilling Reports and Logs, Energy and Geoscience Institute at the University of Utah, Geothermal Data Repository, DOI 10.15121/1814488.
imported_and_pipeline_archive_registered · CC BY 4.0 as displayed on the GDR submission page
McLennan, John (2021), Utah FORGE: Well 78B-32 Daily Drilling Reports and Logs, Energy and Geoscience Institute at the University of Utah, Geothermal Data Repository, DOI 10.15121/1814488.
FORCE 2020 organizers; source logs from Norwegian government; distributed through Zenodo
Competition set from 118 Norwegian Sea wells; logs are described as slightly cleaned and partially despiked. Record metadata says lithofacies/lithology interpretations were hand-crafted by geoscientists. Imported normalized rows and the sampled runtime edition are recorded in the separate well-logs release, not the research SQLite release.
imported_and_pipeline_archive_registered · Mixed/component-specific: Zenodo record declares CC BY 4.0; description identifies source well logs as NLOD 2.0
Bormann, P.; Aursand, P.; Dilib, F.; Manral, S.; Dischington, P. (2020), FORCE 2020 Well well log and lithofacies dataset for machine learning competition, DOI 10.5281/zenodo.4351156. Also cite the FORCE competition as required by its source documentation.
Equinor ASA and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS
Equinor states the release contains about 40,000 files from the Volve field, which produced from 2008 to 2016, including subsurface, operational and production data. Current official access is through Equinor Open Data on Databricks Marketplace; access guide instructs user to sign in or sign up, Get Access, then wait for listing/catalogue availability (up to one hour).
linked_only · Equinor Open Data Licence / Terms and Conditions for Licence to Data – Volve
Credit Equinor and former Volve licence partners (ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS), link the terms, and include copyright notice if applicable.
imported · Equinor Open Data Licence / Terms and Conditions for Licence to Data – Volve
Credit Equinor and former Volve licence partners (ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS), link the terms, and include copyright notice if applicable.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Workbook mirror supplied by Jamiu Olamilekan Badmus via Kaggle; not fetched from Equinor.
Historical Volve workbook from a pinned public third-party mirror, not directly authenticated against Equinor. Source zero/null, flow kind and well type are retained, including inconsistent row labels. Daily columns do not declare most units: these remain unknown; monthly sheet explicitly gives hours and Sm3. Monthly dates retain month precision. Daily and monthly sheets overlap and must not be summed together. No forecast, current status or point-in-time availability is inferred.
imported · Equinor Volve data terms; third-party workbook mirror
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Workbook mirror supplied by Jamiu Olamilekan Badmus via Kaggle; not fetched from Equinor.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Workbook mirror supplied by Jamiu Olamilekan Badmus via Kaggle; not fetched from Equinor.
Historical Volve workbook from a pinned public third-party mirror, not directly authenticated against Equinor. Source zero/null, flow kind and well type are retained, including inconsistent row labels. Daily columns do not declare most units: these remain unknown; monthly sheet explicitly gives hours and Sm3. Monthly dates retain month precision. Daily and monthly sheets overlap and must not be summed together. No forecast, current status or point-in-time availability is inferred.
imported · Equinor Volve data terms; third-party workbook mirror
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Workbook mirror supplied by Jamiu Olamilekan Badmus via Kaggle; not fetched from Equinor.
Contains data under the Norwegian licence for Open Government data (NLOD) distributed by the Norwegian Offshore Directorate.
Source test/engineering summaries, not daily production or current well status. Source QC and update dates are not measurement times; observedAt stays null. Source zeros are retained and flagged as unqualified, not changed to missing. Diameter source strings are retained; decimal and fractional inches are parsed exactly, while uninterpreted labels stay null with a flag. No geographic or physical-well matches are inferred.
Contains data under the Norwegian licence for Open Government data (NLOD) distributed by the Norwegian Offshore Directorate.
Source test/engineering summaries, not daily production or current well status. Source QC and update dates are not measurement times; observedAt stays null. Source zeros are retained and flagged as unqualified, not changed to missing. Diameter source strings are retained; decimal and fractional inches are parsed exactly, while uninterpreted labels stay null with a flag. No geographic or physical-well matches are inferred.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
imported · CC BY-NC-SA 4.0 linked by conversion publisher; no resale
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
imported · CC BY-NC-SA 4.0 linked by conversion publisher; no resale
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Petroware conversion of source LAS/DLIS, not verified against original binaries. Conversion disclaimer and noncommercial/share-alike restrictions apply. Full source JSON downloaded; runtime contains sampled source cells only. Header date is retained as log metadata, not assigned to individual depth measurements. Unknown units/datum remain null.
imported · CC BY-NC-SA 4.0 linked by conversion publisher; no resale
Equinor and former Volve licence partners ExxonMobil Exploration & Production Norway AS and Bayerngas Norge AS. Original data also credit Norwegian Petroleum Directorate; JSON conversion by Petroware AS.
Source: NLOG / TNO Geological Survey of the Netherlands.
Selected pore-fluid pressures underlying overpressure mapping, from activities through 2012. Heterogeneous tests and interpreted values with source reliability/remarks; not current telemetry. Only Pressures_Total imported: stratigraphic sheets overlap and are not double-counted. Depth is TVDSS in m; coordinates remain source UTM31 with datum unverified and are not mapped as WGS84.
Contains information provided by the North Sea Transition Authority and/or other third parties.
Historical disclosed production, not current telemetry. All 182 index entries collected as 181 unique ZIPs (one shared URL retained once); runtime imports sampled original cells from 150 CSV files. The 31 Excel workbooks are imported separately in nsta-ukcs-daily-workbooks-sample with reviewed per-well scope and units. No daily/monthly aggregation, gap filling, unit conversion or physical-well matching. CSV headers and source units retained; Buchan units are explicit in column names.
20,000 synthetic analytical-model responses, 75 columns. Not observed field wells. Full original archived externally; no model trained or benchmark accuracy claimed. Repository code licence is Apache 2.0; CSV dataset licence is separately CC BY 4.0. Structural audit checks finite values, duplicate vectors and seven generator classes. Barrier distance 1,000,000 encodes an absent barrier; it is not a measured physical distance. Generator inputs and response outputs require task-specific leakage controls.
Official package metadata identifies merged QERDB petroleum, QCOS coal, TerraSearch minerals and Groundwater components. Petroleum-only source lineage and component rights remain unverified. The official ZIP returned an AWS WAF challenge (HTTP 202; zero bytes) on 7 October 2026, so no archive tables or normalized rows were inspected or imported.
Australian Government Bioregional Assessment Program
Original archive collected; no measurement rows republished. Manually transcribed Galilee well completion test summaries and derived hydraulic-head sheets, with separate CSIRO PressurePlot material. Pressure units are ambiguous in some sheets; raw and interpreted results overlap.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
Well-level DIGITAL/SCANNED report availability, not actual DST test measurements or downloaded report contents. Event dates, depths and coordinates in the original index have not been promoted to test observations or verified spatial measurements. No current well status or physical-well cross-source matching is inferred. Both availability kinds for the same KID remain distinct.
All rows in one Alberta Public Data monthly volumetric extract, including facility/activity/product reporting rows and well-linked production rows; mixed facility, activity, product and transaction scopes; not conventional-only and not a count of distinct physical wells. Every CSV row is represented. Every populated measurement channel is a separate observation with all 30 native source columns retained. An all-blank row is carried by its real blank Volume cell, flagged no_numeric_measurement_present. No aggregation, conversion, deduplication or physical-well reconciliation is performed. Count of nonblank raw cells in Volume, Energy, Hours, ProrationFactor, and Heat; includes nonnumeric source sentinels which remain raw and have null normalized values.
imported · Government of Alberta non-commercial reproduction permission
Source: Government of Alberta, Alberta Public Data / Petrinex, Alberta volumetric reporting, production month 2025-01. Crown copyright acknowledged. This is a reproduction/adaptation and is not an official version or affiliated with or endorsed by the Government of Alberta, AER or Petrinex.
Contains information provided by the North Sea Transition Authority and/or other third parties.
Historical disclosed daily records from 31 per-well post-COP workbooks. Not current telemetry. Only rows and columns explicitly tied to the workbook well are considered. Platform/field totals, unallocated platform streams and inferred values are excluded. Shell/Heron imports only allocated per-well fluid columns and online hours. Banff platform choke and injected-gas density columns are excluded because their well-level scope or units are unclear. Source dates, raw values and units are preserved; no unit conversion, interpolation or deduplication between source files is performed. Some workbooks publish the same quantity in separate unit columns (for example Sm3 and MMSCF); those columns are alternate representations and must not be summed. Identifiers name source wellbores and are not asserted to represent unique physical wells.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
A bounded sample of eight TestDataSave rows for three source KIDs, not a statewide or complete KGS DST collection. KGS KID/API identifiers are source identifiers, not independently resolved physical-well identities. Bottom-hole temperature comes from CSV rows with explicit DEG F. Pressure summary cells come only from detail-page blocks uniquely matched by KID, test number, date, and time; the page labels PSI and the matched CSV declares PSIg. CSV depth values have no unit field and remain raw metadata only. Source date and time fields are kept separately; observation date has day precision and source time has no documented timezone. Linked vendor PDF is archive-only and excluded from normalized content; pressure curves are omitted.
NLOG / TNO Geological Survey of the Netherlands; cite NLOG as the source and state modifications.
Official NLOG Datacenter well production/injection output, source-supplied wide monthly cells grouped by source well name, field and operator, 2003–2026. The API response has no stable well identifier; the composite source key used for grouping is not a reconciled physical-well identity. A returned zero remains zero; source null remains null; no absent month is filled. Values are per API product and operation, not summed across products. Monthly periods have YYYY-MM precision; future month columns have null observation dates with their source period retained separately. Actual UTC acquisition time is separate from the server Date header and historical availability. Raw JSON does not declare units. UI unit selectors and gas scaling do not authenticate raw response units; normalized units remain unknown pending a matched official data export. Recent months may be provisional or not yet supplied; the provider says monthly data are published after a reporting lag. This is not sensor telemetry or a forecast label.
U.S. Geological Survey, NWIS records retrieved through the USGS/EPA Water Quality Portal; indicate normalization changes.
Historical WQX 2.2 NWIS response for Texas groundwater chemistry. This profile excludes USGS additions after March 11, 2024; requested date bounds do not establish actual coverage. Concentration fractions, analytical methods, detection conditions, quantitation limits, qualifiers and QC samples must be reviewed before training. Non-detects are null, not zero or the quantitation limit. Source agency/site/activity identifiers do not reconcile physical wells. Acquisition is not historical availability. Source clock time and timezone code are retained without inventing a UTC instant.
U.S. Geological Survey, NWIS records retrieved through the USGS/EPA Water Quality Portal; indicate normalization changes.
Only source stations matching the collected chemistry records are normalized; the complete downloaded Texas station CSV remains external. Well/hole depths and site elevation are distinct; source units, horizontal datum, vertical datum, aquifer, well type and source construction-date text are preserved. A construction date is not a depth measurement date or historical availability; unknown observation dates stay null. NAD83 is not asserted to be WGS84, and no coordinates are converted.
Petrobras / 3W Project, dataset v2.0.0; credit the pinned source instance, source licence and normalization changes. No endorsement.
Five deliberately small real instances from five anonymized source-well groups, not the complete or balanced 3W dataset. Simulated/drawn instances are excluded. Naive source clocks have no verified timezone; public observation dates have day precision with exact source clocks in the payload. Source identity is not a reconciled physical well. Label/state roles and codebooks are explicit; exclude target labels from predictors. Provider labels are not predictions, failure onset or universally transferable ground truth. Historical availability, calibration and production model accuracy are not established.
Every listed byte was checked against its pinned Git blob and archived locally. Original network download time was not observed and is unknown. File and anonymized source-group counts are not physical-well counts or measurement-cell counts. Real, simulated and drawn instances remain separate. Only the separately receipted five-real-instance subset is normalized for public runtime. No full-corpus model readiness or validated accuracy is claimed.
this uses Environment Agency flood and river level data from the real-time data API (Beta)
This is the 2,784 source readings returned for one Environment Agency groundwater station; the requested September 1 start was clipped by source retention to September 6. It is not a source-wide coverage claim. Acquisition time is not historical availability. Station IDs do not identify physical wells. Values are groundwater level relative to mAOD, not pumping or flood labels. No forecast or model performance is claimed.
657,012 report rows and 48,080 distinct source API/sidetrack tuples, not physical wells. Official oil/water units BBL and gas MCF are recorded in the dictionary; pressure units are unspecified. Report/revision/formation rows need reconciliation before aggregation. No public normalized values are exposed.
New pressure/temperature/flow source lead. No source bytes were acquired; elevated Pason file download was cancelled by the user after a sandbox error. The cancelled download was not retried. Item metadata and HEAD are not import, unit, time or count verification.
Kuniansky and Bellino, USGS Data Series 669, database updated 2016-04-19
Archived source table and lookup dictionaries are retained externally. Original HTTP receipt, text encoding and compilation rights remain unverified; no normalized scientific observation dataset or ML targets have been imported.
U.S. Geological Survey, Water Data APIs. Cite the monitoring-location and daily-series records; identify normalization changes.
This is a bounded source snapshot, not all USGS groundwater observations. Daily values for parameter 72019 are depth below land surface; only daily mean statistic 00003 series are selected. Native units remain attached. Source last_modified is retained as record refresh metadata, not historical availability. Today’s API response cannot establish when a past observation was first publicly available. Site IDs identify monitoring locations, not reconciled physical wells. Only agency_code=USGS records are included.
Irregular instantaneous dipped measurements in native mAOD, not depth below land surface or daily means. No interpolation or zero filling. Bounded station subset; source GUIDs are not physical-well reconciliation. Source timezone, historical availability and parsed coordinate CRS remain unknown. Original bounded first-1000 station list, sorted source GUIDs, qualified dipped histories spanning at least one year. Original selection probes were not retained; proof covers retained import responses only.
Source: U.S. Bureau of Safety and Environmental Enforcement (BSEE), OGOR-A Production 2025, downloaded directly from the BSEE Data Center.
The live archive is a current annual snapshot, not a preserved historical vintage. Blank volumes are unknown/missing and remain NULL; source zeros remain numeric zero. No physical-well deduplication, aggregation, or joins to separate lease/borehole files are performed. BSEE/BOEM disclaim data accuracy/completeness; public copying language does not establish separate commercial reuse or ML terms for the underlying reporting records.
imported · BSEE website public-information copying permission with requested credit; no separate open license or explicit commercial/ML terms identified for this production record.
Source: U.S. Bureau of Safety and Environmental Enforcement (BSEE), OGOR-A Production 2025, downloaded directly from the BSEE Data Center.
Source: NOAA National Centers for Environmental Information (NCEI), Global Historical Climatology Network Daily (GHCN-Daily).
Bounded deterministic subset of at most 20 contiguous-US stations selected nearest a fixed set of broad regional targets among stations whose inventory spans 2016–2025 for PRCP/TMAX/TMIN; not source-wide coverage. Rows are daily station observations and climate covariates, not well, groundwater, precipitation-at-well, or recharge measurements; no well matching or spatial equivalence is asserted. Native integer values and NOAA element units are retained without conversion: PRCP is tenths of mm; TMAX/TMIN are tenths of degrees Celsius. Missing sentinel -9999 is null with provenance; PRCP measurement flag T is null rather than treated as zero. Quality, measurement, and source flags are retained verbatim as flags and raw records; no quality-flag filtering is applied. Station metadata provides latitude, longitude, and elevation in metres (-999.9 means missing); parsed coordinates are not reprojected and CRS datum is not asserted. Daily source date is retained without asserting a timezone or observation boundary. Historical availableAt is unknown. Retrieval time records this acquisition, not when each historical observation became available; source archive may change. GHCN-D combines multiple source networks and QA-reviewed records; daily dates may follow local/source reporting conventions, not a globally synchronized day. Source documentation disclaims accuracy/reliability/completeness warranties; use requires independent suitability review.
imported · NOAA/NCEI public-domain notice for government information, subject to annotated copyright and third-party exclusions. This is a US-location GHCN-Daily subset; preserve constituent source flags and the documented WMO/non-US restrictions.
Source: NOAA National Centers for Environmental Information (NCEI), Global Historical Climatology Network Daily (GHCN-Daily).
Joint Organisations Data Initiative / International Energy Forum
Complete original oil IVT archive retained; IVT rows were not normalized or counted. A separate 2026 primary CSV has 134,400 year-to-date rows (January–July) and is not the full oil history. Public normalized redistribution rights remain unverified.
Complete original gas CSV archive (2009-01–2026-07), 315,758 country/flow/product/unit reporting rows; raw source codes and assessment flags retained. Public normalized redistribution rights remain unverified.
Source: U.S. Bureau of Safety and Environmental Enforcement (BSEE), OGOR-A annual production archives, downloaded directly from the BSEE Data Center.
Current revised annual snapshots, not historic available-at vintages. Source row grain is completion/month; API sidetracks and completions are not independently verified physical wells. No aggregation, imputation, unit conversion or cross-source matching. 2025 bytes are verified but normalized rows are excluded here to avoid duplicating the separate unchanged bsee-ogora-well-production-2025 dataset. 2026 is incomplete and its last supplied month is reported in coverage. Raw complete records remain in source ZIPs; public measurement payloads preserve exact row pointers and production context.
Source: U.S. Energy Information Administration (EIA), State Energy Data System (SEDS), release 2026-06-26.
Complete allowlisted annual series from the official SEDS bulk source. Geography includes U.S. national and state/DC records plus only the identified federal offshore Gulf of America (X3) and Pacific (X5) production regions. Offshore regions are separate geography entities, not state rows; the U.S. total can include federal offshore production and need not equal the sum of states. State estimates are energy-system statistics, not well-level production. Native annual quantity, period, measure, and unit are retained. Production/consumption estimates may be statistical estimates described in SEDS technical notes. Historical availableAt is unknown; acquisition time is not historical vintage availability.
Source: Statistics Canada, Crude oil and equivalent, monthly supply and disposition, reference dates 1985-01 through 2016-02. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Source: Statistics Canada, Crude oil and equivalent, monthly supply and disposition, reference dates 1985-01 through 2016-02. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Source: Statistics Canada, Supply and disposition of crude oil and equivalent, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Source: Statistics Canada, Supply and disposition of crude oil and equivalent, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Source: Statistics Canada, Natural gas supply and disposition, monthly, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
Values are preserved at native table grain; no conversion of scalar factors or units was performed. Blank VALUE remains SQL NULL; numeric zero remains 0. STATUS, SYMBOL, TERMINATED and DECIMALS are retained per source row. The table reference period is month precision; no day or timezone is assigned.
Source: Statistics Canada, Natural gas supply and disposition, monthly, reference dates 2016-01 through 2026-07. Reproduced and distributed on an ‘as is’ basis with the permission of Statistics Canada.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
KGS source is county and statewide monthly aggregate production and producing-well counts, not per-well production or telemetry. Do not add statewide totals to the county dataset. The current source states data are through May 2026; later monthly cells are retained as source nulls and are not treated as observed. Values are preserved as reported without unit conversion, interpolation or gap filling.
The source of this material is the Kansas Geological Survey website at http://www.kgs.ku.edu/. All Rights Reserved.
KGS source is county and statewide monthly aggregate production and producing-well counts, not per-well production or telemetry. Do not add statewide totals to the county dataset. The current source states data are through May 2026; later monthly cells are retained as source nulls and are not treated as observed. Values are preserved as reported without unit conversion, interpolation or gap filling.
All 56 monthly archives currently listed by the official portal, January 2022–August 2026, are retained privately with SHA-256 receipts: 30,985,924 original mixed facility/activity/product CSV reporting rows. This is not a unique-well count. January 2025 has a separately reviewed normalized/public sample; other months remain original structured CSV archives, not public normalized measurements. September and October 2026 were not listed.
archive_only · Government of Alberta non-commercial reproduction permission
Source: Government of Alberta, Alberta Public Data / Petrinex. Crown copyright acknowledged. Reproduction/adaptation; not an official version or endorsement.
