CEAtlas is built exclusively from open-licensed data plus
Compounding Energy's own proprietary forecast outputs. This page
lists every data source, its licence, and the attribution we owe.
We publish it in full because using someone's data without crediting
them isn't a shortcut worth taking.
The short version. Every map layer is either
open-licensed (OGL v3, CC-BY 4.0, CC0, ODbL, or comparable) or a
Compounding Energy proprietary forecast product. One source —
FracTracker's U.S. Data Centers tracker — is used under
non-commercial-with-credit terms, shown free to every user and never
behind a paid tier. Details for each are below.
Infrastructure
Layer
Source
Licence
Transmission lines
PyPSA-Eur (EU cleaned grid: 220 kV and up in the PyPSA-Eur countries other than Great Britain; gridkit is the 2016 ENTSO-E-map extraction PyPSA-Eur uses) · OpenStreetMap (global, and the primary source in the United States, where the underground HV cables of 100 kV and up are served with the overhead lines from the first nightly build after 2026-10-13; 33–219 kV lines in the PyPSA-Eur countries) · United States, from the same build: HIFLD Open Electric Power Transmission Lines (the final public issue, frozen 2025-08-26, as archived by SeerAI on Source Cooperative), drawn only where no line the map already shows has the line (the OpenStreetMap lines of this layer and of the Offshore cables layer), at every voltage — each such line says so in its popup; HIFLD's file is the United States', and two of its lines run on about 4 km into Canada. Great Britain, from the first nightly build after 2026-10-03: OpenStreetMap lines at 33–219 kV and cables at 100–219 kV from our own GB extract (a way — an OpenStreetMap line feature — carrying both a ≥ 220 kV and a sub-220 kV voltage is shown at its sub-220 kV level; busbars, bays and out-of-service ways are left out); from the first nightly build after 2026-10-12, its cables at 33–99 kV on land too (a cable under water — offshore-wind inter-array and island cabling — is left out). Great Britain at 220 kV and up, from the first nightly build after the GB backbone switch of 2026-10-04: OpenStreetMap lines and cables from the same extract (the 220 kV offshore-wind export cables included; a way carrying two voltages is then shown at both), Great Britain's internal HVDC links and the four interconnectors gridkit lacks (Viking, IFA2, ElecLink, Greenlink), each as whole OpenStreetMap ways; PyPSA-Eur gridkit keeps Northern Ireland, Ireland and the continent, the cross-border HVDC links it has (BritNed, IFA, Nemo including its Great Britain half, Moyle, EWIC, NSL) and eleven 275 kV circuits (London, Sheffield, Glasgow and Edinburgh city routes and three Whitelee wind-farm connections, 92 km) that OpenStreetMap does not map continuously at 275 kV. Until that build, and on any night the GB extract is missing or substantially short, 220 kV and up in Great Britain is PyPSA-Eur gridkit. Queries and rules: OpenStreetMap-derived data: method and offer
ODbL (OSM) · MIT/CC-BY (PyPSA-Eur) · CC BY 4.0 (the SeerAI archive of HIFLD Open; HIFLD Open is US public data). Attribution: “HIFLD Open (DHS), archived by SeerAI, CC BY 4.0”. The first build to serve them was the nightly of 2026-10-13 (11,584 HIFLD gap-fill lines, about 76,900 km); before it the served layer carried none (from 2026-08-29). HIFLD lines are also used in the Electricity border crossings build (below) and the CONUS engine networks (the proprietary section).
Substations
PyPSA-Eur · OpenStreetMap (incl. 75k US via chunked Overpass queries — Overpass is OpenStreetMap's query service; every GB substation)
ODbL (OSM) · MIT/CC-BY (PyPSA-Eur)
GB wind turbines (OpenStreetMap) — a separate layer, UK regional cut (served since 2026-10-03: the first tile, built that day from a fetch with the widened sea boxes, went out with the deploy; the nightly's gb_wind group refreshes it monthly)
OpenStreetMap only: power=generator + generator:source=wind elements (a way at its centre) and power=plant wind relations / ways (plant:source~wind) in the United Kingdom (land and territorial waters) plus the offshore sea boxes listed on the method page; nothing from any other source is mixed into this layer, and no OpenStreetMap value enters the REPD plant register, the fleet census, the BM-unit table or the GB wind validation
OpenStreetMap (transmission + distribution; worldwide from the first deploy after 2026-11-03: every man_made=pipeline way of the gas family — substance gas, natural_gas, cng, lpg, ngl, hydrocarbons — read through osmium from the 14 Geofabrik continent extracts, Europe and the United States of 2026-11-02 and the other continents of 2026-10-16, 108,566 ways; before that build, 8 Overpass boxes that left out south-east Asia, India, East Asia below 18 N, Alaska west of 140 W, Hawaii and Central America) · U.S. EIA Natural Gas Interstate and Intrastate Pipelines (January 2020 shapefile)
OpenStreetMap (waterway=dam, man_made water-supply tags, water and sewage pipelines) worldwide, from the Geofabrik continent extracts of 2026-11-02 (every continent but Antarctica; Greenland's own extract not fetched) — built 2026-11-03; Europe + North America Overpass boxes before that
OpenStreetMap compressor stations (substation=compression) via the Overpass API · U.S. EPA Greenhouse Gas Reporting Program RY2023 data summary (transmission-compression and storage facilities, subpart W)
U.S. EIA natural gas processing plants (EIA-757 Schedule A, 2017) · EPA GHGRP RY2023 (processing facilities) · North American Cooperation on Energy Information (NACEI) processing plants, Canada and Mexico (2016–17)
US public domain (EIA, EPA) · Open Government Licence – Canada (NACEI). Attribution: “Source: U.S. Energy Information Administration (EIA)”; “U.S. EPA Greenhouse Gas Reporting Program, RY2023”; “Contains information licensed under the Open Government Licence – Canada; North American Cooperation on Energy Information (NRCan/EIA/SENER)”.
Gas facilities — underground storage
U.S. EIA underground storage fields (Dec 2020 points) with EIA-191 2024 field capacity via EIA’s Natural Gas Annual Respondent Query System · HIFLD Open Natural Gas Underground Storage (Dec 2023), archived by SeerAI on Source Cooperative (locations for fields missing from the 2020 file) · NACEI underground storage, Canada (2016)
US public domain (EIA) · CC BY 4.0 (SeerAI archive of HIFLD Open) · Open Government Licence – Canada (NACEI). Attribution: “Source: U.S. EIA (EIA-191 via Natural Gas Annual Respondent Query System)”; “HIFLD Open (DHS), archived by SeerAI, CC BY 4.0”; the NACEI sentence above.
Gas facilities — LNG terminals
U.S. EIA, U.S. Liquefaction Capacity (2026 Q2) · U.S. EIA LNG import/export terminals (Apr 2020) · HIFLD Open LNG Import and Export Terminals (Aug 2024, archived by SeerAI) · NACEI LNG terminals, Canada and Mexico (2017) · site locations from a list we keep, one documented source per site: EIA, EPA GHGRP, EPA Facility Registry Service, Wikidata or OpenStreetMap
Gas facilities — border crossings (gas and liquids)
U.S. EIA natural gas and liquids pipeline border crossings (Nov 2017; the US side) · NACEI border crossings (the Canadian and Mexican side’s own reports, 2017; merged with the US point within 5 km)
US public domain (EIA) · Open Government Licence – Canada (NACEI)
Gas facilities — trading hubs
U.S. EIA natural gas trading hubs (Feb 2020)
US public domain. Attribution: “Source: U.S. Energy Information Administration (EIA)”.
Gas pipeline projects
U.S. EIA, Natural Gas Pipeline Projects tracker (Aug 2026) · geometry from OpenStreetMap gas pipeline ways matched by name
U.S. EPA Greenhouse Gas Reporting Program, RY2023 data summary spreadsheets (direct emitters in industry; reported as of 2024-08-16). Emissions only — not electricity demand.
US public domain. Attribution: “U.S. EPA Greenhouse Gas Reporting Program, RY2023”.
EIA’s Energy Atlas feature services now require a token, so these layers are built from EIA’s static map-data
files and tabular releases; the Esri Federal re-hosts (Esri Master License Agreement) are not used. EPA GHGRP is
frozen at reporting year 2023. Canadian and Mexican facilities come from the 2016–17 NACEI release.
Mines, fuel geology, geothermal & earth imagery
Layer
Source
Licence
Mines (coal, metal & minerals)
Mine Safety and Health Administration (MSHA) Mines data set (weekly) · U.S. EIA, EIA-7A Annual Coal Production (2024), joined on the MSHA id · Natural Resources Canada, Principal Mineral Areas, Producing Mines, and Oil and Gas Fields (900A), 2025 (metal, non-metal, coal and oil-sands mines)
US public domain (MSHA, EIA) · Open Government Licence – Canada (NRCan). Attribution: “Mine Safety and Health Administration (MSHA) Mines data set; Source: U.S. EIA, EIA-7A Annual Coal Production (2024)”; “Contains information licensed under the Open Government Licence – Canada (Natural Resources Canada, Principal Mineral Areas, Producing Mines, and Oil and Gas Fields (900A), 2025)”.
Coal fields
U.S. Geological Survey Open-File Report 96-92, Coal Fields of the Conterminous United States (Tully, 1996), with Alaska coalfield points · basin names from the U.S. EIA sedimentary basins (May 2011)
US public domain. Attribution: “U.S. Geological Survey Open-File Report 96-92 (Tully, 1996)”; “Source: U.S. Energy Information Administration (EIA)”.
Oil & gas plays and fields
U.S. EIA tight oil and shale gas plays, Lower 48 (Dec 2021) and sedimentary basins (May 2011) · NRCan 900A principal producing oil and gas fields, Canada (2025)
US public domain (EIA) · Open Government Licence – Canada (NRCan). Attribution: “Source: U.S. Energy Information Administration (EIA)”; the NRCan sentence above.
Geothermal heat (3 km)
Stanford Thermal Earth Model for the Conterminous United States, Geothermal Data Repository submission 1592: the 3 km temperature, its uncertainty and the modelled heat flow, re-gridded to 0.03° cloud-optimised GeoTIFFs
CC BY 4.0. Attribution: “Aljubran, M., & Horne, R. (2024). Stanford Thermal Earth Model for the Conterminous United States [Dataset]. Geothermal Data Repository. https://doi.org/10.15121/2324793”.
NASA daily satellite imagery · NASA Night basemap
NASA Global Imagery Browse Services (GIBS) WMTS, gibs.earthdata.nasa.gov: VIIRS (NOAA-20, NOAA-21, Suomi NPP) and MODIS (Terra, Aqua) corrected-reflectance true colour, one day at a time; VIIRS City Lights 2012 for the NASA Night basemap. Tiles are loaded by your browser from NASA; nothing is stored.
NASA imagery, no restrictions on use. Acknowledgement: “We acknowledge the use of imagery provided by services from NASA's Global Imagery Browse Services (GIBS), part of NASA's ESDIS.”
Day / Night (live)
Computed in the browser from the sun’s position (standard astronomical formulae) — no third-party data
—
MSHA publishes its mines file weekly; the popup shows the file’s date. The USGS coal-field map (1996) and the EIA play
outlines (2021) are static compilations, not resource or reserve estimates. The geothermal layer is a model, not a
measurement. US field-level oil and gas footprints are not shown yet.
Scottish offshore wind — the location of REPD offshore projects with no usable REPD coordinates (lease-area centroids; Ossian, MarramWind, MachairWind, Morven …) and, as an evidenced in-range override, of projects whose REPD coordinates lie outside their own lease (Beatrice, Moray West); the same point places the Beatrice census farm (Moray West's census farm sits at its DUKES point, 0.02 km from the lease centroid). A lease-derived point is never on-array evidence for a BM unit: the Beatrice BM units are held off the GB BM Actions map pending review. Contains public sector information, licensed under the Open Government Licence v3.0, from Crown Estate Scotland.
Open Government Licence v3
The Crown Estate — Wind Site Agreements (England, Wales & NI)
English / Welsh offshore wind — the location of REPD offshore projects with no usable REPD coordinates (lease-area centroids; Norfolk Boreas / Vanguard, Dogger Bank South, Morgan, Morecambe, North Falls) and, as an evidenced in-range override, of projects whose REPD coordinates lie outside their own lease (Triton Knoll, East Anglia One); the same two points place those census farms. Contains data provided by The Crown Estate that is protected by copyright and database rights.
A turbine grid reference stated in the planning authority's own published committee report for the farm's application (Lambs Hill: Stockton-on-Tees Borough Council, application 10/2549/EIS, turbine T2) — a single published fact, cited to the report; no open licence is claimed over the report itself.
Cited public record
Operator-published site positions
The owner / operator's own published site position where every official register files the farm elsewhere (Minsca: Nadara Ltd, 'Our sites' asset map) — a single factual coordinate copied as published and cited; no open licence is claimed, and the BM-unit table labels such a unit 'cited-fact' rather than OGL.
Cited published fact
DUKES 5.11 (Digest of UK Energy Statistics, DESNZ)
GB fleet census — major-producer operational fleet (113.9 GW) plus the dated closure ledger (38.1 GW); the 2019–2025 editions as published (stations operational at the end of May of each year, fetched from the Internet Archive's captures of the gov.uk files) give a metered site's capacity for a year where no dated per-unit value passes (the capacity-factor basis of the GB metered sites)
Open Government Licence v3
Elexon BMRS / Insights — B1610 settlement-metered generation (monthly, and half-hourly for the capacity checks) · reference/bmunits/all (unit names, fuel types, lead parties) · REMIT normal capacity · IGCPU installed capacity · MELS maximum export limits
CECadence — GB BESS open feed (the Elexon BM registrations of GB battery units: unit, name, MW, active in the latest quarter, postcode; the feed the GB BESS layer draws)
ENTSO-E Transparency terms (free re-use with attribution)
JRC-PPDB-OPEN (JRC Open Power Plants Database) — historical
Fed the EU fleet census until 29 Aug 2026, when the census was rebased onto GEM GIPT (a frame change: JRC's 645 GW large-unit, 2019-snapshot fleet → GEM's utility-scale tracker). No live layer reads it today; credited for the provenance of earlier snapshots. Zenodo DOI 10.5281/zenodo.3574566
World fleet census backbone — August 2026 snapshot: 30,093 plants ≥ 20 MW (7,940.4 GW operating / 612.5 GW retired / 5,119.5 GW pipeline; Europe on its own GEM tab, March 2026 release), plus rest-of-world plant detail on the map
French plant register on the map — from the first nightly build after 2026-10-06 it defines French onshore wind and solar (PowerPlantMatching and GEM operating rows for those fuels are dropped); installation-level capacity, status and metered rolling-12-month energy; points are commune centroids unless the row below places them
Positions of Hauts-de-France wind farms on the map, from the first nightly build after 2026-10-06 — an ODRÉ wind row matched to a park in this built-turbine register is drawn at the park's turbine centroid (location_precision: turbine_park_centroid; each such row names this source and the park's update date in location_source and, from the first nightly build after 2026-10-07, carries the licence's attribution — producer, dataset, Licence Ouverte v2.0 and the date of the dataset's last update, one statement for the whole edition — in location_attribution, which CSV exports credit); unmatched rows stay at the commune centroid
Meta · Google · Microsoft · xAI · OpenAI / Oracle public pages
Public disclosures, attributed
FracTracker U.S. Data Centers
The FracTracker U.S. Data Centers layer (the only
open community-opposition dataset we know of) comes from the
FracTracker Alliance,
a 501(c)(3) non-profit. Records are compiled from public permit
documents, media monitoring, and community submissions, verified
before publication.
It is used with attribution under FracTracker's
non-commercial-with-credit terms — "downloaded and shared
freely for non-commercial use with credit to FracTracker." CEAtlas
displays this layer free to every user, never behind any
paid tier. Every popup carries the line "Data: FracTracker
Alliance" and a link to their tracker. If FracTracker requests
removal, we take the layer down within 24 hours.
OpenStreetMap, worldwide — every underwater power cable it maps (offshore-wind export and inter-array cables, island links, HVDC interconnectors) and overhead power lines marked as a shipping seamark; from Geofabrik's continent extracts (Europe and the United States as of 2026-11-01, the other continents as of 2026-10-15; built 2026-11-03; until then the eight Overpass boxes of the 2026-04 build). Queries, rules and the free machine-readable copy: method and offer (§10)
OpenStreetMap (terrestrial fibre and telecom cable ways) worldwide, from the Geofabrik continent extracts of 2026-11-02 (every continent but Antarctica; Greenland's own extract not fetched) — built 2026-11-03; 7 European Overpass boxes before that
NESO Open Data Licence — “Supported by National Energy SO Open Data”
GB BM Actions
Elexon settlement acceptance volumes (BOAV) + acceptance flags (BOALF). Unit locations: Wikidata (BM-unit id P11610 → coordinate) and the DESNZ Renewable Energy Planning Database (via Wikidata’s REPD id). From the next GB BM Actions rebuild after 2026-10-06 (the layer is rebuilt weekly, not by the deploy): where our location corrections moved a unit’s REPD record, the unit sits at that source’s point and carries its licence and statement — Crown Estate Scotland, HES Canmore, a DUKES OS reference, an operator’s published position (a cited fact) — see the Power plants rows above; each unit’s popup and CSV row names its basis and statement, and the tile attribution appends every statement present.
NYISO MIS public CSV archive (damlbmp / rtlbmp) · MISO market reports (da_expost_lmp / rt_lmp) · ISO-NE histRpts (WW_DALMP / rt-lmp) · PJM Data Miner 2 (public anonymous key) · SPP Marketplace portal (da-lmp-by-location / rtbm-lmp-by-location) · CAISO OASIS (PRC_LMP / PRC_INTVL_LMP) · ERCOT DAM via gridstatus.io (keyed) or the CompoundVision price-store export · EIA-930 net generation by fuel and interchange (the fuel-share and seam-consistency checks)
Public ISO/RTO data under each market's own terms · gridstatus.io terms (ERCOT) · US public domain (EIA-930). Shown beside our forecasts and used to score and calibrate them: the scorecard charts the posted hourly hub and zone prices next to the model's, and the chart files it links carry those values (up to the latest 14 days hourly; daily figures in each region's history). The feeds themselves are not redistributed — there is no feed download, only the values the comparison uses.
Electricity prices — EU
ENTSO-E Transparency Platform A44 day-ahead (41 bidding zones, drawn as 27 country polygons — IT / NO / SE / DK volume-weighted)
Country electricity mix (Country Mix layer; the CSV export's country rows)
Ember / Our World in Data energy data (/data/country_energy.json: generation mix, demand, low-carbon and renewables shares by country) · country outlines Natural Earth 110m
CC BY 4.0 (Ember, Our World in Data) · public domain (Natural Earth). Attribution: “Ember / Our World in Data energy data (CC BY 4.0); country outlines Natural Earth (public domain)”.
EU generation mix
ENTSO-E Transparency Platform A75 (actual generation per production type)
Each ISO's public queue file — CAISO public queue report · NYISO interconnection queue · MISO GI queue · ISO-NE IRTT public queue · SPP GI summary · ERCOT GIS Report · PJM planning queue (CAISO, NYISO, MISO and ISO-NE land today) · county centres from the U.S. Census Bureau 2024 Gazetteer
Respective ISO/RTO terms of use (not an open licence) · US public domain (Census Bureau). Attribution: “Source: CAISO public queue report”, and likewise per ISO (“Source: NYISO interconnection queue”, “Source: MISO GI queue”, “Source: ISO-NE IRTT public queue”, “Source: SPP GI summary”, “Source: ERCOT GIS Report”, “Source: PJM planning queue”); “U.S. Census Bureau, 2024 Gazetteer files”.
HIFLD Electric Retail Service Territories — Oak Ridge National Laboratory / DHS HIFLD Open (Feb-2025 issue, mirrored by Public Environmental Data Partners “HIFLD Next”) · U.S. EIA Form EIA-861 (2024 final; Service_Territory county lists used to check every polygon's location) · U.S. Census Bureau cartographic boundary files 2024 (counties 1:5M, places 1:500k: the outlines that replace misplaced HIFLD polygons) · Ontario LDC service-area boundaries (data.ontario.ca) · Natural Earth province / country outlines (approximate areas for QC, MB, SK, NB, NS, PE, NL, Mexico)
US public data, no use restriction (HIFLD Open; mirror licence field “other”) · US public domain (EIA) · US public domain (Census Bureau) · Open Government Licence – Ontario · public domain (Natural Earth). Attribution: “Oak Ridge National Laboratory / DHS HIFLD Open (via Public Environmental Data Partners)”; “U.S. Energy Information Administration, Form EIA-861 (2024)”; “U.S. Census Bureau, cartographic boundary files (2024)”; “Contains information licensed under the Open Government Licence – Ontario”.
Gas utility territories
HIFLD Natural Gas LDC Service Territories — Oak Ridge National Laboratory / DHS HIFLD Open (2017 issue, via Public Environmental Data Partners) · U.S. EIA Form EIA-176 (2024) via EIA’s Natural Gas Annual Respondent Query System · U.S. Census Bureau 2024 counties (the county split of HIFLD’s single National Grid New York outline) · Ontario natural gas distributor service-territory boundaries (data.ontario.ca)
US public data (HIFLD Open) · US public domain (EIA) · US public domain (Census Bureau) · Open Government Licence – Ontario. Attribution: “Oak Ridge National Laboratory / DHS HIFLD Open (via Public Environmental Data Partners)”; “U.S. Energy Information Administration, Form EIA-176 (2024)”; “U.S. Census Bureau, cartographic boundary files (2024)”; “Contains information licensed under the Open Government Licence – Ontario”.
Balancing authorities · ISO / RTO regions
HIFLD Control Areas (Oak Ridge National Laboratory / DHS HIFLD Open, modified 2025-04-08, via Public Environmental Data Partners) re-keyed to the U.S. EIA-930 BA register (2026) and clipped to their interconnection; ISO/RTO footprints and the SWPW / BHBA / SIKE areas derived by Compounding Energy; Canadian and Mexican areas (and New Brunswick, added to NBSO) from Natural Earth outlines
US public data (HIFLD Open) · US public domain (EIA-930) · public domain (Natural Earth)
Reliability coordinators
HIFLD NERC Reliability Coordinators (DHS HIFLD Open, modified 2025-07-07, via Public Environmental Data Partners); IESO, Hydro-Québec and New Brunswick areas from Natural Earth outlines
US public data (HIFLD Open) · public domain (Natural Earth)
Interconnections
Derived by Compounding Energy from HIFLD NERC Regions (DHS HIFLD Open via Public Environmental Data Partners: WECC → Western, Texas RE → ERCOT, the rest → Eastern; Black Hills’ area → Western per EIA-861) and Natural Earth province / state outlines
US public data (HIFLD Open) · public domain (Natural Earth)
FERC regions
HIFLD FERC Regions — Federal Energy Regulatory Commission / DHS HIFLD Open (modified 2024-04-25, via Public Environmental Data Partners)
US public data (HIFLD Open)
HIFLD Open closed on 2025-08-26; these boundaries are the last public issue, as preserved by
Public Environmental Data Partners
(an archived copy is also published by SeerAI on Source Cooperative under CC BY 4.0). They are frozen and drift
from today's utilities; the EIA-861 / EIA-176 numbers are refreshed each year.
Planned transmission & predicted grid
Layer
Source
Licence
US Planned Transmission (≥ 230 kV)
Our Grid Future (Horizon Energy Systems + Great Plains Institute aggregation of US utility / RTO / TSO filings)
Underlying data public-domain via DOE / FERC / EIA; OGF aggregation used with attribution
WECC path ratings (Scenario Studio region-border limits)
Western Electricity Coordinating Council (WECC), 2026 Path Rating Catalog, public version — the published path-rating figures (the accepted import / export ratings of rated WECC paths) to be used for path-rating figures where a Scenario Studio region border crosses a rated path
WECC terms — wecc.org/legal-and-privacy: WECC's materials “may be viewed, downloaded, copied or reproduced only in compliance with applicable law”. Figures only, used with attribution: “based on published WECC path ratings (2026 Path Rating Catalog, public version)”; the catalogue itself is not redistributed.
The predicted grid (GridFinder) is credited under
Infrastructure; GridFinder lines tagged
source=osm in the original dataset are excluded to avoid
double-counting our existing OSM ingest.
Generation pipeline & grid edge
Layer
Source
Licence
Wind turbines
United States Wind Turbine Database v9.0 (June 26, 2026) — U.S. Geological Survey, American Clean Power Association and Lawrence Berkeley National Laboratory · Canadian Wind Turbine Database — Natural Resources Canada (updated 2026-04-24) · Mexico: OpenStreetMap wind generators (power=generator + generator:source=wind) and their power=plant wind farms, from the Geofabrik Mexico extract of 2026-11-02 (served from the 2026-11-03 rebuild; each Mexican feature carries src OSM, and the layer is free — OpenStreetMap data sits only in free layers)
United States Large-Scale Solar Photovoltaic Database v4.0 (April 2026) — U.S. Geological Survey and Lawrence Berkeley National Laboratory
US public domain (USGS data release). Attribution: “Fujita, K.S., Ancona, Z.H., Kramer, L.A., et al., 2026, United States Large-Scale Solar Photovoltaic Database (ver. 4.0, April 2026): USGS and LBNL data release, https://doi.org/10.5066/P9IA3TUS”.
Electricity border crossings
Border Crossings: Electric Transmission Line — North American Cooperation on Energy Information (NRCan / EIA / SENER; 2017) · HIFLD Electric Power Transmission Lines (the North American lines in our 2026-08-18 lines file, which the 2026-09-29 build read — a static build, not rebuilt by the nightly) intersected with the U.S. Census Bureau 2024 cartographic boundary states (1:500k; Natural Earth selects the Canadian / Mexican stretches) · U.S. Department of Energy, Office of Electricity, Presidential Permit Library
Open Government Licence – Canada · US public data (HIFLD Open) · US public domain (Census Bureau, DOE) · public domain (Natural Earth). Attribution: “Contains information licensed under the Open Government Licence – Canada; North American Cooperation on Energy Information (NRCan/EIA/SENER)”; “U.S. Department of Energy, Office of Electricity, Presidential Permits”; “U.S. Census Bureau, cartographic boundary files (2024)”.
Plant changes (EIA-860M) · planned generators on the Power Plants layer
Form EIA-860M Preliminary Monthly Electric Generator Inventory — two consecutive monthly vintages (Planned sheet: planned operation month) — U.S. Energy Information Administration; change classification by CEAtlas · plants EIA lists without coordinates placed at U.S. Census Bureau 2024 Gazetteer county centres
US public domain (EIA asks for source attribution). Attribution: “U.S. Energy Information Administration, Form EIA-860M Preliminary Monthly Electric Generator Inventory (<Month YYYY>); change classification by CEAtlas.”
Not included, by licence: Mexican wind turbines and Canadian / Mexican solar arrays exist openly only in
OpenStreetMap, and mixing ODbL data into these mixed-source layers would raise the share-alike question for the
whole layer, so they stop at the sources above. Where we do serve OpenStreetMap turbines — Great Britain, the GB wind turbines layer served since 2026-10-03 — it
is a separate, all-OpenStreetMap layer under the ODbL (the GB wind turbines
row; queries and rules on the method and offer page).
Nothing on this map is taken from other commercial map products.
Capacity-factor & resource data
Layer
Source
Licence
Hourly weather store (86-year record)
ERA5 reanalysis, 1940–2025, Copernicus Climate Change Service (C3S) — powers the CF maps, 24/7 CFE and connection studies, drought and exceedance products
Open-Meteo archive API (ERA5-family hourly weather where the CE store is not used), commercial API plan
CC-BY 4.0
7-day production forecast (plant popups) & the Data Assistant's wind/solar outlook
Open-Meteo forecast API, commercial API plan — GFS Seamless for the forecast panel; best-match models for the assistant, which at UK locations can include UK Met Office UKV forecasts
CC-BY 4.0 (Open-Meteo); Met Office data CC BY-SA 4.0
GB zonal & European price-forecast layers (CEGridSight)
Open-Meteo forecast API, commercial API plan — weather inputs processed by Compounding Energy (CEGridSight) into the price forecasts; GB weather from the UK Met Office models
CC-BY 4.0 (Open-Meteo); Met Office data CC BY-SA 4.0
Open-Meteo forecast and archive APIs, commercial API plan — US wind/solar weather and temperatures (US national weather models), processed by Compounding Energy into the capacity factors and demand inputs of the daily price solves
CC-BY 4.0
Wind capacity factor
DTU Global Wind Atlas v3 (cf_iec2, native 250 m/pixel)
CC-BY 4.0 (ECMWF open data terms; modified products attributed)
Solar capacity factor
ESMAP / World Bank Global Solar Atlas v2 (PVOUT)
CC-BY 4.0
Site Value — the ~1 km resource behind the indicative LCOE
The CEAtlas 86-year ERA5 capacity-factor record (P50 / P90), bilinear from 0.25°, × the Global Wind Atlas 3 (DTU / World Bank Group, ESMAP; IEC class 2 capacity factor, 250 m) and Global Solar Atlas 2 (Solargis / World Bank Group, ESMAP; PVOUT, ~1 km) textures, block-averaged to ~1 km worldwide; capacity factors capped at 70 % (wind) and 43 % (solar)
CC-BY 4.0 (Global Wind Atlas, Global Solar Atlas); Copernicus licence (ERA5)
The Global Wind Atlas is released by DTU and the World Bank; the Global
Solar Atlas by ESMAP / the World Bank Group. Both are available under
CC-BY 4.0
and are attributed accordingly.
Copernicus / ERA5 attribution. Contains modified
Copernicus Climate Change Service information (1940-2025). Neither the
European Commission nor ECMWF is responsible for any use that may be
made of the Copernicus information or data it contains.
Open-Meteo attribution.Weather
data by Open-Meteo.com, used under
CC BY 4.0
(Open-Meteo licence)
on Open-Meteo's commercial API plan. The data have been changed: Compounding Energy
processes them into capacity-factor, generation and study estimates (the CE weather
store's ERA5 hourly record is built from Open-Meteo's open-data copy of ERA5; the
point-study fallback, the 7-day production forecast and the assistant's outlook call the
Open-Meteo API). Weather-derived API answers carry the same credit in an
X-Data-Attribution header and a Link header
(rel="license").
Our capacity-factor chain is validated against independent references —
and those references have owners too. Methods and results are published
in full at /methodology.
Source
Used for
Licence
PVGIS (European Commission, Joint Research Centre)
Solar CF chain validation — hourly seriescalc cross-checks at globally-stratified sites
Natural Earth Data (urban areas, protected areas) · OpenStreetMap (airports)
Public domain (Natural Earth) · ODbL (OSM)
Exclusion zones — the route finder's constraints in the United States, Great Britain and Europe (route_constraints.pmtiles, drawn inside the same free layer)
The route builds' own designations and marine protected areas: the Route finder rows below (PAD-US; Natura 2000 and NatDA; Natural England, NatureScot, Scottish Government, NRW, DAERA; NOAA MPA Inventory).
As those rows: public domain (PAD-US, NOAA CC0) · CC BY 4.0 (EEA) · OGL v3.0 (UK agencies).
Exclusion zones — the global route finder's constraints outside those countries (route_constraints_global.pmtiles, drawn inside the same free layer, from 2026-11-03)
OpenStreetMap, via Geofabrik's continent extracts (2026-10-16): the protected areas, nature reserves and national parks (boundary=protected_area / national_park, leisure=nature_reserve, classed no-go or sensitive by protect_class), military land (landuse=military, military=*) and aerodromes (aeroway=aerodrome areas; a 1.5 km disc round an aerodrome point with an IATA / ICAO code) — classified exactly as the global 2 km route grids classify them, and left out inside the countries whose designations come from the route builds above. Countries from Natural Earth (the rule only; no Natural Earth geometry is served in this file). Method: OpenStreetMap-derived data: method and offer.
Exclusion zones — Planning ban (wind): Scotland's two National Parks and 40 National Scenic Areas (friction.pmtiles, drawn inside the same free layer, from 2026-11-03); the same file carries the local planning authority outlines the GB Planning Risk Index layer draws
NatureScot / Scottish Government National Park and National Scenic Area boundaries (as the route build merged them); the rule: Scottish Government, National Planning Framework 4, Policy 11(b), adopted 13 February 2023 — "Development proposals for wind farms in National Parks and National Scenic Areas will not be supported." Authority outlines: ONS Local Planning Authorities (January 2026) Boundaries UK BUC (V2).
Exclusion zones — US planning friction tier: one strictness tier (rules recorded / setback or siting rules / restrictive limits / reported ban or moratorium, or preempted by state law) per US county or township × technology (us_friction.pmtiles, drawn inside the same free layer, from 2026-11-03). An area rule summary from AI-assisted collection, not legal advice; bans are reported, not yet checked.
National Renewable Energy Laboratory (NREL), U.S. Wind Siting Regulation and Zoning Ordinances (2025), OEDI submission 8519, doi:10.25984/3363758 (data 2025-09-12) and U.S. Solar Siting Regulation and Zoning Ordinances (2025), OEDI submission 8602, doi:10.25984/3363739 (data 2025-12-30) — collected with generative AI (INFRA-COMPASS): "The Large Language Models used for this effort make mistakes. Always validate the data for critical use cases." The tier, its thresholds and the state-law gate are Compounding Energy's reading. Boundaries: the jurisdiction polygons NREL ships in its reVX GeoPackages, under the same CC BY 4.0; measured on 2026-11-03 against the U.S. Census Bureau 2024 cartographic boundary files (cb_2024_us_county_500k, held locally), the 932 county rings follow the Census county line — 93.3 % of their 244,376 vertices lie within 15 m of it, 98.9 % within 100 m — but are not that file (the vertex sets differ; the widest departures are shoreline and river counties); the township / town / city rings were not measured.
CC BY 4.0 (attribution: National Renewable Energy Laboratory, U.S. Wind and Solar Siting Regulation and Zoning Ordinances, 2025 — the rows and the polygons) · the county rings follow the Census 2024 county line as measured (the Census files themselves are US public domain)
Planning Risk Index (GB, indicative) — the per-authority refusal index and the planning record, by size, consent route, nation and period (friction_index.pmtiles; Analyst, Great Britain). The map draws council outlines only
DESNZ Renewable Energy Planning Database (REPD), Q2 2026 extract — statuses and stage dates per application, aggregated per authority × technology × era and, for the record, per technology × size × route × nation × period (no project, applicant or operator is named). Tested against, and feeding no figure: the Planning Inspectorate's Casework Database (October 2026 file, England; names dropped at read). A note on court challenges: the independent review into legal challenges against nationally significant infrastructure projects (published 28 October 2024). This layer's figures no longer use DESNZ, Electricity Generation Costs 2025, or NREL ATB 2025's battery capex: here they fed only the £ planning contingency shown from 3 to 4 November 2026, which was withdrawn on 4 November 2026 and removed from the data file (NREL ATB 2025 still feeds Site Value; see its row).
OGL v3.0 (DESNZ; Planning Inspectorate; the review) · Boundaries as the row above
Power of Place-West — environmental exclusion categories (Cat 1–3, wind + solar, 11 western US states) — free layer (United States group)
The Nature Conservancy — Wu, G.C. et al. (2023), Minimizing habitat conflicts in meeting net-zero energy targets in the western United States, PNAS; data release v2, Zenodo doi:10.5281/zenodo.11268042. Shown as schematic scenario-analysis data per TNC’s use limitation (not a siting determination).
CC BY 4.0 (attribution: The Nature Conservancy / Wu et al. 2023)
Protected Areas Database of the United States (PAD-US) 4.1, Federal Management Agencies — U.S. Geological Survey Gap Analysis Project (USGS GAP ArcGIS Online service)
US public domain. Attribution: “U.S. Geological Survey (USGS) Gap Analysis Project (GAP), 2024, Protected Areas Database of the United States (PAD-US) 4.1: USGS data release, https://doi.org/10.5066/P96WBCHS”.
BLM Solar Zones & Energy Corridors
BLM Energy Designations (EGIS): Solar Energy Zones, Section 368 designated corridors (current alignment + centerlines) — Bureau of Land Management, National Operations Center
US public domain. Attribution: “Bureau of Land Management, National Operations Center; West-wide Energy Corridors (Section 368, EPAct 2005)”.
Offshore Wind Leases (BOEM)
BOEM Renewable Energy GIS — Offshore Wind Lease Outlines; Offshore Wind Planning Area Outlines (rescinded 30 July 2025) — Bureau of Ocean Energy Management
US public domain. Attribution: “Bureau of Ocean Energy Management (BOEM), Office of Renewable Energy Programs”.
Natural Hazard Risk (FEMA NRI)
FEMA National Risk Index v1.20 (December 2025), counties — FEMA’s National Risk Index county feature service (services.arcgis.com/XG15cJAlne2vxtgt … National_Risk_Index_Counties/FeatureServer/0), accessed 2026-09-29 · county outlines: U.S. Census Bureau cartographic boundary files (2024)
OpenFEMA terms of use (US federal data; no FEMA logos). Attribution: “FEMA National Risk Index v1.20 (Dec 2025). This product uses the Federal Emergency Management Agency's OpenFEMA API, but is not endorsed by FEMA.”
Wildfire Hazard Potential (USFS 2023)
Wildfire Hazard Potential for the United States (270-m), version 2023, classified — USDA Forest Service (Interagency Image Processing Platform image service)
USDA Forest Service data — can be used without additional permissions or fees. Attribution: “Dillon, G.K. 2023. Wildfire Hazard Potential for the United States (270-m), version 2023. USDA Forest Service RDS-2015-0047-4”.
Population Density (GHSL 2025)
GHS-POP R2023A, epoch 2025, 30 arc-second grid — European Commission Joint Research Centre (converted to people per km² by CEAtlas)
U.S. Census Bureau cartographic boundary file, counties 1:500,000 (2024) · Statistics Canada, 2021 Census – Boundary files, census divisions (cartographic) · INEGI, Marco Geoestadístico 2024 (áreas geoestadísticas municipales)
US public domain · Statistics Canada Open Licence · Términos de Libre Uso de la Información del INEGI. Attribution: “U.S. Census Bureau, cartographic boundary files (2024)”; “Adapted from Statistics Canada, 2021 Census – Boundary files (census divisions, cartographic), 2022. This does not constitute an endorsement by Statistics Canada of this product.”; “Fuente: INEGI, Marco Geoestadístico 2024 (áreas geoestadísticas municipales). Simplified by Compounding Energy for web display; not an official INEGI product.”
The hazard layers are relative screening indices, not site-specific probabilities, and none of these sources endorses
CEAtlas. The WECC CEM inputs layer (Desk) draws land cover derived from Esri / Impact Observatory 10 m Annual Land Use
Land Cover (CC BY 4.0) and slope from USGS GMTED2010 (public domain). Nothing on this map is taken from other
commercial map products.
Land values (official averages)
Layer
Source
Licence
Land Values (USDA + FHFA)
USDA National Agricultural Statistics Service — Census of Agriculture 2022 (farmland value per acre, land + buildings) and the Land Values / Cash Rents surveys, from the Quick Stats bulk files (qs.census2022, qs.economics) · FHFA — The Price of Residential Land for Counties, ZIP Codes, and Census Tracts in the United States, v4.0 (June 2024; FHFA Working Paper 19-01) · county outlines: U.S. Census Bureau cartographic boundary files (2024)
US public domain. Attribution: “USDA NASS”; “This product uses FHFA data but is neither endorsed nor certified by FHFA.”
Land Values — England (MHCLG)
Ministry of Housing, Communities and Local Government — Land value estimates for policy appraisal 2023 (valuation date 1 October 2023) · Office for National Statistics — Local Authority Districts (December 2025) and ITL1 (January 2025) boundaries, ultra-generalised
Eurostat — Agricultural land prices by region (apri_lprc) and Agricultural land rents by region (apri_lrnt), NUTS 2 / NUTS 1 / national · region outlines: Eurostat GISCO NUTS 2024, 1:3 million
These layers show published official averages for whole areas. They are not valuations or appraisals of any site and
none of these publishers endorses CEAtlas.
Flood & water risk
Layer
Source
Licence
Flood Zones — Great Britain (EA · NRW · SEPA)
Environment Agency — Flood Map for Planning (Rivers and Sea), Flood Zones 2 and 3 · Natural Resources Wales — Flood Map for Planning, Flood Zones 2 and 3 (rivers and sea, merged) · Scottish Environment Protection Agency — Flood Maps v3.0, river and coastal flood extents (high / medium / low likelihood)
European Commission Joint Research Centre, Copernicus Emergency Management Service — European river flood hazard maps v3.1.1 (EFAS; doi:10.2905/1D128B6C-A4EE-4858-9E34-6210707F3C81) and global river flood hazard maps v2.1.2 (GloFAS)
World Resources Institute — Aqueduct Floods v2, coastal inundation, historical, 1-in-100-year, no subsidence
CC BY 4.0. Attribution: “WRI Aqueduct Floods”.
Water Risk (WRI Aqueduct 4.0)
World Resources Institute — Aqueduct 4.0 water risk atlas (Kuzma, S., M.F.P. Bierkens, S. Lakshman, T. Luo, L. Saccoccia, E.H. Sutanudjaja and R. van Beek. 2023. Aqueduct 4.0: Updated decision-relevant global water risk indicators. Technical Note. WRI); basin outlines from the Aqueduct 4.0 database (HydroBASINS level 6)
CC BY 4.0. Attribution: “WRI Aqueduct 4.0”.
Water Scarcity — Europe (EEA WEI+)
European Environment Agency — Water Exploitation Index plus (WEI+) at sub-unit level, 2000–2023, v01 (doi:10.2909/b16bd284-f2ec-4164-90b7-674c1de399ba)
U.S. Drought Monitor — weekly map (National Drought Mitigation Center, University of Nebraska-Lincoln; USDA; NOAA)
Free use with credit. Attribution: “The U.S. Drought Monitor is jointly produced by the National Drought Mitigation Center at the University of Nebraska-Lincoln, the United States Department of Agriculture, and the National Oceanic and Atmospheric Administration. Map courtesy of NDMC.”
SEPA Flood Maps v3.0 — the source attribution statements SEPA requires with its data (river and coastal)
These are modelled or planning-scale screening layers, not site flood risk assessments or water-availability
decisions, and none of these publishers endorses CEAtlas.
Connection costs (the cost book)
Used for
Source
Licence
Connection Study costs — Europe, GB, NI (and elsewhere, transferred)
Source: ACER (European Union Agency for the Cooperation of Energy Regulators), Energy infrastructure unit investment cost (UIC) indicators, 24 April 2026 — https://www.acer.europa.eu/sites/default/files/documents/Publications/ACER-2026-Unit-Investment-Cost-Indicators.pdf (accessed October 2026); GB / NI converted at the ECB euro reference rates, 2025 annual average; the local price level and labour factor as on Site Development Cost (sources in that row)
ACER reuse policy: reproduction and distribution permitted for non-commercial and commercial purposes, provided ACER is acknowledged as the source.
Connection Study costs — United States
NREL reVX, least-cost transmission configuration (line, substation and transformer costs; land-use and slope multipliers; the line-capacity → voltage table), https://github.com/NatLabRockies/reVX; escalated with the U.S. Bureau of Labor Statistics Producer Price Index WPUIP231232 (net inputs to power and communication structures; via FRED); cross-checked against EIA / Sargent & Lundy, Capital Cost and Performance Characteristics (AEO2025); the local labour factor as on Site Development Cost (sources in that row)
reVX: BSD 3-Clause License — “Copyright (c) 2026, Alliance for Energy Innovation, All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. * Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.” BLS and EIA: public domain (U.S. Government works).
Site Value — indicative LCOE
Compiled by Compounding Energy. Equipment capex and fixed O&M: National Renewable Energy Laboratory (NREL), Annual Technology Baseline 2025 v1.0.0 (Moderate scenario, R&D cost drivers; land-based wind class 4, utility-scale PV class 4; 2023 US$), via the Scenario Studio cost book (7 % real, 30 years). Site works: the Site Development Cost layer (sources in the row below). Resource: the CEAtlas 86-year ERA5 capacity-factor record, sharpened to ~1 km everywhere with the Global Wind Atlas 3 (DTU / World Bank Group, ESMAP) and Global Solar Atlas 2 (Solargis / World Bank Group, ESMAP) downscaling textures (the resource row); capacity factors capped at 70 % (wind) and 43 % (solar) where the resource data's peak is implausible for a 50 MW plant. Price levels: the World Bank's World Development Indicators (PPP conversion factor, GDP, LCU per international $ — PA.NUS.PPP; official exchange rate — PA.NUS.FCRF; latest year, mostly 2024; a value published on the wrong currency basis is corrected (Bulgaria), a territory without its own value takes its parent country's or neighbour's (Liechtenstein: Switzerland; the Isle of Man, Jersey, Guernsey and Gibraltar: the UK; Monaco: France; Åland: Finland; Vatican City: Italy), and a value outside 0.15–1.5, or none, takes the median of the country's World Bank region, or the world median where the region has none). Rest of the world (connection only): the CEAtlas grid (October 2026) — AC lines of 110 kV up to 800 kV and substations of 110 kV and above, HVDC links and converter stations excluded — at the cost book's transferred European (ACER) unit costs (edition 2026-11a). In every region, lakes and permanent ice (Natural Earth, public domain) are left blank. Drawn as one map across every region (one US$ colour scale; euros and pounds at the ECB 2025 rate, for colour only). The Connection view hatches places with no grid nearby: in the US, GB and Europe, places from which the site works reach no grid of the connection voltage over land; elsewhere, places with no tappable ≥ 110 kV line or substation within ~300 km. The figures are shown on the Analyst plan (a relative preview without figures is free); the land adder is given only where the land model has a value (not, for example, in Turkey, Ukraine, Iceland or Cyprus).
Compiled by Compounding Energy from the cost book above (NREL reVX, ACER UIC 2026, EIA AEO2025 location factors, BLS PPI), TxDOT average low bid unit prices and the FHWA National Highway Construction Cost Index, BLS OEWS May 2025 (state and metropolitan / nonmetropolitan area files), ONS ASHE 2025, Eurostat (nama_10r_2coe, nama_10r_2emhrw, prc_ppp_ind, lc_lci_lev), UFS 3-701-01 (DoD Facilities Pricing Guide, July 2026: the labour share), PyPSA-Eur / GridKit lines (continental Europe), OpenStreetMap power lines, substations, highways and airports (the points and the outlines behind the no-go bits), HIFLD Open transmission lines filling the US gaps the map draws (from the 2026-10-13 rebuild), U.S. Census TIGER/Line 2025 roads, OS Open Roads, the Copernicus DEM GLO-90, USGS NLCD and Impact Observatory land cover; the land component from the Indicative Farmland Value surfaces
DoD area cost factors (Alaska, Hawaii, Puerto Rico site cost)
UFS 3-701-01, DoD Facilities Pricing Guide, 31 July 2026 (formerly UFC 3-701-01), Table 4-1 MILCON area cost factors, from the 2026 Data Tables workbook (updated 2026-08-04) — Whole Building Design Guide (National Institute of Building Sciences) for the Department of Defense, https://www.wbdg.org/dod/ufc/ufs-3-701-01. The factor of each county, borough or municipio scales the contiguous-US unit costs of the Site Development Cost, Grid Connection by Size and civil works layers in Alaska, Hawaii and Puerto Rico. Census cartographic county boundaries 2024 (cb_2024, U.S. Census Bureau) place them.
A work of the United States Government — “APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED” (the guide’s cover); public domain. Census boundaries: public domain.
Grid Connection by Size (indicative)
Compiled by Compounding Energy: the Site Development Cost connection works at each project size (5–1,000 MW) and the cost book's network reinforcement per MW (NREL US Utility-Scale PV and Land-based Wind Supply Curves 2024, reV; Lopez et al. 2025 — transferred to Europe and GB); spare capacity from the line ratings of the CEAtlas transmission data and, in SSEN's area of Great Britain, Scottish and Southern Electricity Networks' Headroom Dashboard (2026-03), as also used by the Connection Study
Civil works adders: boreholes, landfill, mining, ground hazards
Layer
Source
Licence
Boreholes & ground investigations
British Geological Survey, Single Onshore Borehole Index (SOBI, July 2026); USGS Water Data API monitoring locations (groundwater wells); Basisregistratie Ondergrond (BRO) via PDOK — TNO: cone penetration tests (CPT), geological (BHR-G) and geotechnical (BHR-GT) boreholes; BRGM Banque du Sous-Sol (per-département exports, October 2026); Geological Survey Ireland: geotechnical boreholes, bedrock boreholes (verified and unverified), groundwater wells and springs; Geoscience Australia Borehole Database (web service, November 2026); New Zealand Petroleum and Minerals Geodata Catalogue wells and drill holes (November 2026).
Environment Agency: Historic Landfill Sites and Permitted Waste Sites — Authorised Landfill Site Boundaries (site outlines, references, dates, waste types, gas / leachate controls; operator, licence-holder and site names and addresses are not carried); NIEA authorised landfill sites 2017 (OpenDataNI); US EPA Landfill Methane Outreach Program (LMOP) Landfill and Project Database (September 2024); US EPA Superfund National Priorities List sites with status, National NPL Superfund Site Boundaries, and Facility Registry Service interests (SEMS, ACRES brownfields); European Environment Agency, Industrial Reporting under the Industrial Emissions Directive (site map; E-PRTR Annex I activity 5(d) landfills, the latest reporting year per site).
The Coal Authority: Development High Risk Area and Development Low Risk Area (the coal mining reporting area); Mine Safety and Health Administration (MSHA) Mines data set (operators not carried); OSMRE Abandoned Mine Land Inventory System (e-AMLIS) problem areas; USGS Mineral Resources Data System (MRDS; producers and past producers in the United States); Natural Resources Canada, Principal Mineral Areas, Producing Mines, and Oil and Gas Fields (Map 900A, 75th edition, 2025); Gouvernement du Québec, SIGEOM — Indices, gîtes, mines et carrières (November 2026).
Coal Authority: Open Government Licence v3.0 — “Contains public sector information from The Coal Authority licensed under the Open Government Licence v3.0”. MSHA, OSMRE, USGS MRDS: public domain (U.S. Government works). Natural Resources Canada: Open Government Licence – Canada. Gouvernement du Québec (SIGEOM): Creative Commons Attribution 4.0 (CC BY 4.0) — “Source: Gouvernement du Québec, SIGEOM”.
Ground hazards & peat
BGS GeoSure 5 km hex grid v8 (shrink-swell, landslides, soluble rocks, compressible ground, collapsible deposits, running sand); USGS, Karst in the United States: a digital map compilation and database (Weary & Doctor 2014, Open-File Report 2014-1156); BRGM / Géorisques, exposition au retrait-gonflement des argiles (2021); NatureScot, Carbon and Peatland 2016 map — Wind Farm Spatial Framework (Class 1 and 2 only); PEATMAP (Xu, Morris, Liu & Holden 2018, Research Data Leeds, doi:10.5518/252).
Compiled by Compounding Energy (edition civil-2026-10b; the transmission line view's tower spacing, foundation cost and premiums are stated assumptions): earthworks unit prices from TxDOT statewide average low bid unit prices (items 110 excavation and 132 embankment, twelve-month moving, August 2024) × the FHWA National Highway Construction Cost Index, moved by the labour and civil-engineering price levels of the Site Development Cost layer (BLS, ONS, Eurostat); USDA NRCS gNATSGO (July 2020) map units via the Microsoft Planetary Computer (depth to bedrock, annual minimum water table, hydrologic group, Vertisol and Histosol major components); the four layers above; slope and flood shares as Site Development Cost (Copernicus DEM GLO-90; Copernicus EMS / JRC river flood hazard). Quantities, premiums and contingencies are stated assumptions (grade C). The parcel tool reads the same rasters.
gNATSGO: Creative Commons CC0 1.0 (USDA NRCS). TxDOT, FHWA: public record / public domain. The other sources: as their rows above and in the cost book section.
Route finder: least-cost line routes
Input
Source
Licence
Costs per km
Compiled by Compounding Energy (edition route-2026-11c): the Site Development Cost layer's line costs, terrain and land-cover multipliers and connection points (see the cost book), the civil works adders per km of line, and an easement from the modelled land values. The corridor widths, the easement share and the sensitive-area multipliers (× 2.5 overhead, × 1.25 for a buried cable) are stated assumptions.
As their source rows on this page.
HVDC lines and cables, converter stations
HVDC overhead line per km as a share of the top AC class line (0.827 ×) and the US converter station costs ($150k LCC / $195k VSC per MW, the MISO 2025 Transmission Cost Estimation Guide's figures): NREL ReEDS-2.0, inputs/transmission/rev_transmission_basecost.csv and inputs/scalars.csv. DC cable per km (submarine DC, P25 / median / P75), AC submarine cable per km (an overhead AC line's water crossings) and European converter stations per MW: ACER, Unit Investment Cost Indicators 2026 (Annex I). The AC cable in the contiguous US is ACER's price × 4, an estimate, as we found no public per-km price for a 345 kV AC submarine cable there: ISO New England, 2050 Transmission Study (2024); American Transmission Co., Straits of Mackinac cable replacement (2021, as reported by the Associated Press); Engineering News-Record, Long Island Replacement Cable (2008); Prysmian, Elba to Piombino contract (2021); NYISO, Long Island Public Policy Transmission Need fact sheet, and New York DPS, Propel NY Energy project notice (a total, no cable price). The HVDC cable in the contiguous US is ACER's price × 2.6, an estimate: ISO New England, 2050 Transmission Study (2024), Table 5-2, its HVDC cable cost from NREL's Atlantic Offshore Wind Transmission Study, less the 50 % contingency the study added. Parsons Brinckerhoff / IET, Electricity Transmission Costing Study (2012): a cross-check of the cable cost, and the significant-crossing allowance (a crossing of a motorway, railway, navigable waterway or another line ≈ 0.7–0.9 km of the line's cost, from its crossing sensitivity); the crossings themselves from CEAtlas's transport rights-of-way and transmission layers and OpenStreetMap. A converter station's local part (civil works and buildings, erection and commissioning, engineering, about a third, 32 %): the typical cost structure in Rudervall, Charpentier and Sharma, High Voltage Direct Current (HVDC) Transmission Systems Technology Review Paper, Energy Week 2000 (ABB and the World Bank). Compiled in /data/route_line_types.json.
Water depth (HVDC cable crossings of sea and lakes)
NOAA National Centers for Environmental Information, ETOPO 2022 Global Relief Model, 30 arc-second surface elevation (doi:10.25921/fd45-gt74), sampled at each 1 km cell; a lake's depth measured from its own surface level.
US public domain — “NOAA National Centers for Environmental Information. 2022: ETOPO 2022 Global Relief Model (30 arc-second edition). doi:10.25921/fd45-gt74”.
Protected and sensitive areas — United States
USGS Gap Analysis Project, Protected Areas Database of the United States (PAD-US) 4.1: Protection Status by GAP Status Code (GAP 1 and 2) and Federal Management Agencies (military land); Natural Earth parks.
PAD-US: public domain — “U.S. Geological Survey (USGS) Gap Analysis Project (GAP), 2024, Protected Areas Database of the United States (PAD-US) 4.1”. Natural Earth: public domain.
Protected and sensitive areas — Europe
Natura 2000 (end-2023 release) and the Nationally designated areas inventory (NatDA v24), via the European Environment Agency.
Protected and sensitive areas — Great Britain and Northern Ireland
Natural England (SSSI, SAC, SPA, Ramsar, National Nature Reserves, National Parks, National Landscapes); NatureScot (SSSI, SAC, SPA, National Nature Reserves); Scottish Government (National Parks, National Scenic Areas); Natural Resources Wales via DataMapWales (SSSI, SAC, SPA, Ramsar, AONB, National Parks, National Nature Reserves); DAERA / NIEA via OpenDataNI and the DAERA open data hub (ASSI, AONB, SAC, SPA).
Marine protected areas (avoided by HVDC cables; drawn in Exclusion Zones)
United States: NOAA Office of National Marine Sanctuaries, Marine Protected Areas Inventory 2023. Great Britain and Northern Ireland: Natural England and JNCC, Marine Conservation Zones (England); NatureScot, Nature Conservation Marine Protected Areas (Scotland); Natural Resources Wales via DataMapWales, Marine Conservation Zone; DAERA / NIEA, Marine Protected Areas in the Northern Ireland marine plan area. Marine Natura 2000 sites as in the Europe row above.
The global 2 km grids and the 5 km world grid (outside the contiguous US, Great Britain and Europe, the lines that leave them, and a long line whose 100 m search runs out of time)
Country price levels: World Bank, International Comparison Program 2021 (construction and GDP price level indexes). Land cover: Copernicus Global Land Service, Land Cover 100 m collection 3, epoch 2019 (doi:10.5281/zenodo.3939050). Terrain: NOAA ETOPO 2022 (above). Protected areas and military land: OpenStreetMap, via Geofabrik's extracts, except in Alaska, Hawaii, Puerto Rico and the US Virgin Islands, which keep the US designations (PAD-US: the row “Protected and sensitive areas — United States” above). Airports: OpenStreetMap, via Geofabrik's extracts and, in the United States (Alaska and Hawaii included), Puerto Rico, the US Virgin Islands, Great Britain and Europe, the OpenStreetMap airport points of the Exclusion Zones layer and the outlines of the airports OpenStreetMap maps as areas, via Geofabrik's US and Europe extracts (2026-11-01) and its Central America extract (2026-10-15). Countries: Natural Earth.
Copernicus DEM GLO-90 (slope per 90 m pixel); USGS Annual NLCD 2025 (30 m, United States); Impact Observatory / Esri 10 m annual land use and land cover (Europe); airports from OpenStreetMap.
Transport rights of way: highways, railways, navigable waterways
Used for
Source
Licence
Highways & Major Roads — United States (National Highway System)
FHWA National Highway System, NHS Version 2025.08.08, via the Bureau of Transportation Statistics National Transportation Atlas Database (NTAD). Metadata entry: https://www.arcgis.com/home/item.html?id=dce9f09392eb474c8ad8e6a78416279b
A work of the United States Government; the NHS database, or any portion of it, may be freely distributed as long as its metadata entry is included — linked here and on the layer card.
Highways & Major Roads — United States (measured cross-sections)
FHWA Highway Performance Monitoring System (HPMS) 2024, national full-extent sections (HPMS2024.gdb, data.transportation.gov / BTS item 5e6a977c2d7c4ec1bdc82e684d3384f2): lane width, through lanes, median type and width, shoulder type and widths, widening potential on the Sample Panel; item definitions per the HPMS Field Manual (December 2016)
Public domain (U.S. Government work).
Highways & Major Roads — United States (roads off the NHS)
U.S. Census Bureau, TIGER/Line Shapefiles 2025, Primary and Secondary Roads (PRISECROADS), https://www2.census.gov/geo/tiger/TIGER2025/PRISECROADS/
Federal Railroad Administration (FRA) / Bureau of Transportation Statistics (BTS) [distributor], NTAD North American Rail Network Lines (July 2026); railroad names from the FRA Railroad, Company and Organization List (data.transportation.gov rx2s-7tqf)
A work of the United States Government, available for unrestricted public use (NARN); public domain (the FRA list).
Navigable Waterways — United States
U.S. Army Corps of Engineers / Bureau of Transportation Statistics, NTAD National Waterway Network (2025)
A federal (USACE / BTS) NTAD product, treated as a work of the United States Government; the item carries no separate licence text.
Navigable Waterways — lock points (United States)
USACE Navigation Data Center / Bureau of Transportation Statistics [distributor], NTAD Locks
A work of the United States Government, available for unrestricted public use.
Navigable Waterways — lock points and the Canal & River Trust waterways (England & Wales)
Canal & River Trust, Locks (lock register, weekly; https://www.arcgis.com/home/item.html?id=345dda94d2c6412cae3eb6a1d00d162c). The Trust’s canal centrelines are NOT used (their licence excludes commercial use).
Open Government Licence v3.0 — “Contains Canal & River Trust data licensed under the Open Government Licence v3.0.”
Highways & Major Roads (every country but the United States and Great Britain), Railways (every country but the United States, Canada and Mexico) and Navigable Waterways (every country but the United States) — worldwide since 2026-11-03
OpenStreetMap via the Geofabrik continent extracts (Europe of 2026-11-01; Africa, Asia, Australia-Oceania, Central America, South America, Canada, Mexico and Russia of 2026-10-15): highway, railway (incl. the disused / abandoned lifecycle tags) and waterway ways, with the navigation tags of their waterway relations
U.S. Department of Energy, National Renewable Energy Laboratory (now the National Laboratory of the Rockies), Alternative Fuels Data Center station locator (electric), via the developer API, fetched 2026-10-08
Public domain (U.S. Government work).
EV Chargers — the 75 countries served from Open Charge Map (incl. Great Britain, Spain, Italy, France, Sweden, Finland, Ireland, Australia)
Open Charge Map API (openchargemap.org), fetched 2026-10-09; each record names its data provider
EV Chargers — the countries served from OpenStreetMap (incl. Germany, the Netherlands, Norway, Belgium, Austria, Denmark, Switzerland)
OpenStreetMap amenity=charging_station (Geofabrik continent extracts of 2026-10-08); a country is served from one source only — never merged with Open Charge Map records
Indicative corridor widths and European vessel classes
Design standards and statute cited in each read-out; UNECE European Agreement on Main Inland Waterways of International Importance (AGN) classification
Cited as facts (widths in metres) with the source named; no text reproduced.
Land use & farmland quality
Layer
Source
Licence
Indicative Farmland Value — U.S. / GB / Europe (modelled)
Modelled by Compounding Energy from the official averages above (USDA NASS; MHCLG; Eurostat), Agreste – SSP / FNSafer valeur vénale des terres 2025, and Destatis 61521-0020 (2025), and the covariates: Natura 2000 (EEA), USGS NLCD, USDA NASS CDL, USDA NRCS gNATSGO (via Microsoft Planetary Computer), Impact Observatory land cover, the GB farmland-quality maps, JRC GHSL population and settlement layers, the Copernicus EMS river-flood maps, PAD-US, the Copernicus DEM GLO-90, and the PLACES Lab 2010 fair-market land value estimates (Nolte 2020, PNAS, doi:10.1073/pnas.2012865117; data doi:10.5061/dryad.np5hqbzq9; CC0) as a model input
Compiled by Compounding Energy from MHCLG Land value estimates for policy appraisal 2023 (residential), HM Land Registry Price Paid Data 2025, the ONS postcode → OA / LSOA / MSOA / LAD best-fit lookup (May 2026) and MSOA (2021) population-weighted centroids; residential cells from Impact Observatory land cover and JRC GHS-POP
Indicative Residential Land Value — U.S. (modelled)
Modelled by Compounding Energy from FHFA, The Price of Residential Land for Counties, ZIP Codes, and Census Tracts in the United States, v4.0 (June 2024; Davis, Larson, Oliner & Shui, FHFA Working Paper 19-01), placed with the U.S. Census Bureau 2010 Gazetteer census-tract points, and the covariates of the farmland surface above (USGS NLCD, JRC GHSL, Copernicus DEM GLO-90, Copernicus EMS river flood, PAD-US, gNATSGO, the PLACES Lab 2010 estimates)
Impact Observatory, Microsoft, and Esri — 10 m Annual Land Use / Land Cover (v02), 2024; resampled by CEAtlas to ≈ 100 m worldwide (majority class) and kept at 10 m for Great Britain and Northern Ireland
Creative Commons Attribution 4.0 (CC BY 4.0). Attribution: “Impact Observatory, Microsoft, and Esri”.
Land Cover — U.S. (NLCD 2025)
U.S. Geological Survey — Annual National Land Cover Database (NLCD) Collection 1.2, land cover 2025 (30 m)
U.S. Government work — public domain. Credit: “U.S. Geological Survey, Annual NLCD Collection 1.2 (2025 land cover)”.
Crop Types — U.S. (Cropland Data Layer 2025)
USDA National Agricultural Statistics Service — Cropland Data Layer 2025 (30 m); crop groups by CEAtlas from the CDL class names
Public domain and free to redistribute (USDA NASS). Citation: “USDA National Agricultural Statistics Service Cropland Data Layer. 2025. Published crop-specific data layer. USDA-NASS, Washington, DC.”
Farmland Quality — Great Britain (ALC · LCA)
Defra — Predictive Agricultural Land Classification Map for England (2025) · Natural England — Agricultural Land Classification post-1988 survey polygons · Welsh Government — Predictive Agricultural Land Classification Map 2 · The James Hutton Institute — Land Capability for Agriculture in Scotland 1:250,000
The following layers are Compounding Energy proprietary
products — model outputs, not third-party open data. They are
available to paid subscribers (Enterprise unless stated below) and are not
redistributed under an open licence.
CEGridSight
— European 10-day price / generation / cross-border-flow / carbon
forecasts from a multi-zone unit-commitment MILP.
CompoundVision
— 16-day per-asset wind & solar production forecasts (ensemble).
CENovaSage CONUS day-ahead price forecast
— the nodal run overlay (CENovaSage Run layer; a Desk-tier analysis
feature, with the run list itself free) is a Compounding Energy model output.
Its accuracy record is deliberately public:
/scorecard and the free Price Forecast
Accuracy (CONUS) map layer. The CONUS nodal networks behind the runs are built from the
HIFLD Open Electric Power Transmission Lines file (US public data). The HIFLD lines are used
there, in the Electricity border crossings build (its row above) and in the WECC CEM screening's
transmission-line setbacks, and — from the first nightly build after 2026-10-13 — the map's
Transmission lines layer draws them where no served line has the line (its row above).
CEAtlas proprietary scores — the Connection
Difficulty Score (CDS) and Site Suitability Score are computed by
CEAtlas. Their formulas, inputs, and known limitations are published
at /methodology.
Licence summaries
Licence
What it permits (in brief)
OGL v3
UK Open Government Licence v3.0 — copy, publish, adapt, and exploit (incl. commercially) with attribution. Used for DUKES, REPD and Crown Estate Scotland data.
The Crown Estate Open Data Licence (GIS) v1.1
The Crown Estate's open-data terms — access, adapt, combine and publish its data in the ordinary course of business with the attribution "Contains data provided by The Crown Estate that is protected by copyright and database rights"; no resale of the data itself, no direct commercial gain from supplying it, no use in a competing data portal, no logos or implied endorsement. CEAtlas uses one derived point per offshore project (the lease-area centroid) to place REPD records without usable coordinates and, as an evidenced in-range override, two whose REPD coordinates lie outside their own lease (Triton Knoll, East Anglia One; the same points place those census farms); every such record carries the attribution statement.
CC-BY 4.0
Creative Commons Attribution 4.0 — share and adapt for any purpose, including commercially, with credit. Used for GridFinder, Epoch AI, GEM GIPT, WRI GPPD, Open-Meteo, Global Wind Atlas, Global Solar Atlas, the SeerAI archive of HIFLD Open (LNG terminals, gas storage), JRC GHS-POP (population density), Esri / Impact Observatory land cover (WECC CEM inputs) and the Stanford Thermal Earth Model (geothermal heat).
Copernicus
Copernicus C3S licence — free use of Copernicus Climate Change Service information with the required "Contains modified Copernicus Climate Change Service information" statement, and the disclaimer that neither the European Commission nor ECMWF is responsible for any use made of it. Used for the ERA5 hourly weather store (1940-2025).
NESO's open data licence — re-use, including commercially, with the attribution “Supported by National Energy SO Open Data”. Used for the GB Constraint Boundaries layer (ETYS boundaries + Day-Ahead Constraint Flows and Limits) and for the DNO Licence Areas and GSP Regions layers (NESO's GIS boundary files).
ODbL
Open Database Licence — share, create, and adapt with attribution and share-alike on derived databases. Used for all OpenStreetMap-derived layers and PNNL IM3. The OpenStreetMap-derived layers, their queries and processing rules, and the free-copy offer: /osm-data.html.
OGL – Canada
Open Government Licence – Canada 2.0 — copy, modify, publish and exploit (incl. commercially) with the attribution statement “Contains information licensed under the Open Government Licence – Canada”. Used for the NACEI (NRCan/EIA/SENER) gas facilities and border crossings, and NRCan Map 900A (Canadian mines and oil and gas fields).
Statistics Canada Open Licence
Statistics Canada’s open licence — use, adapt and redistribute (incl. commercially) with the “Source: Statistics Canada …” or, for an adapted product, “Adapted from Statistics Canada … This does not constitute an endorsement by Statistics Canada of this product.” notice. Used for the Canadian census divisions in County Borders.
INEGI free-use terms
Términos de Libre Uso de la Información del INEGI — copy, distribute, adapt and exploit commercially with the credit “Fuente: INEGI, <product>”, telling users about any transformation and never implying INEGI endorsement. Used for the Mexican municipios in County Borders.
OpenFEMA terms
FEMA’s OpenFEMA terms — free use of FEMA’s public data, with the required statement “This product uses the Federal Emergency Management Agency's OpenFEMA API, but is not endorsed by FEMA.”, the dataset and access date cited, and no FEMA logos. Used for the National Risk Index.
CC0
Public-domain dedication — no rights reserved, no attribution required (we credit anyway). Used for Wikidata (including three LNG site locations, the GB census BMU crosswalk's P11610 unit ids and the offshore census sites placed by a Wikidata P625 point).
US public domain
US-government works (EIA, EPA, FCC, HIFLD, NETL, BTS/FRA, NASA POWER, NASA GIBS imagery, MSHA, USGS USWTDB, PAD-US and OFR 96-92 coal fields, the Census Bureau, BLM, BOEM, the USDA Forest Service, the Bonneville Power Administration — BPA's ATCID and OATT postings, to be used for the transmission charges and path limits on the Scenario Studio's region borders) carry no copyright and are free to re-use.
ENTSO-E
ENTSO-E Transparency Platform terms — free re-use of published market data with attribution.
NC-with-credit
FracTracker's non-commercial-with-credit terms — shared freely for non-commercial use with credit. The FracTracker layer is shown free to every CEAtlas user, never behind a paid tier.