Free spacecraft and PV research APIs#

Eight connectors extend the orbit, irradiance and flight-telemetry sources. They discover technical evidence, mission products and geometry for the satellite/component catalogue and subsequent power analysis. They are included in the base orbital-pv installation and need no additional dependencies.

Access and scope#

Available adapters#

Public class

Access

Data and interpretation

NtrsConnector

Anonymous

NASA report metadata and available documents. PV specifications must be extracted and reviewed; some records contain metadata only.

DiscosConnector

Free account/token

ESA object mass, dimensions, identity and mission/launch relationships. No ephemerides or standardized PV-capacity field.

ZenodoConnector

Anonymous public records

Published datasets, DOI/version/license metadata and open file downloads. Restricted files are not unlocked by this connector.

DataverseConnector

Anonymous public datasets

Dataset search, published versions and original file downloads. Defaults to the University of Stuttgart’s DaRUS installation.

HorizonsConnector

Anonymous

JPL vectors for supported spacecraft and other targets, including Sun-relative geometry. These are ephemerides, not measured PV power.

PdsConnector

Anonymous

NASA PDS mission product discovery, labels and data links. Archive coverage and engineering-product availability vary.

PsaConnector

Anonymous public products

ESA PSA ADQL queries with column metadata and download links. Scientific and housekeeping products have different meanings.

DonkiConnector

Anonymous CCMC endpoint

Space-weather event records and links. These provide context, not satellite-specific radiation dose or proven degradation causes.

Seven anonymous services were exercised with real requests on 3 November 2026. DISCOS was tested with mocked responses because no user token was available. Access checks are snapshots, not uptime or permanent access guarantees. Free access does not give every dataset the same redistribution license; retain the record’s license, copyright and attribution requirements.

Response and caching contract#

Every query returns ResearchResponse:

  • data: the original provider JSON structure, preserving pagination, identifiers, units, quality fields, licenses and version metadata;

  • receipt: source, response URL, retrieval time, SHA-256 and byte count;

  • request_url: the exact query URL, without authentication headers;

  • documentation: the provider’s interface documentation.

cache_path writes the original response and a .provenance.json sidecar. Reusing that path verifies both the query and checksum. A different query, missing sidecar or corrupted file raises an error. Use refresh=True to explicitly replace a snapshot. Caches do not expire automatically; inspect receipt.retrieved_at when freshness matters. HTTP failures, including 401/403/429, are raised rather than returned as empty datasets. The original HTTP response exposes Retry-After to the caller; there is no automatic retry loop. Pagination is explicit and bounded to one requested page.

Discover PV specifications in NASA reports#

import orbitalpv as opv

with opv.NtrsConnector() as api:
    page = api.search(
        '"solar array" "end of life"', offset=0, page_size=25,
        cache_path="data/research/ntrs-page-0.json",
    )
    for record in page.data["results"]:
        print(record["id"], record["title"], record.get("downloadsAvailable"))
    report = api.citation(20200000669)
    receipt = api.download_document(
        20200000669, "20200000669.pdf", "data/research/20200000669.pdf",
    )

NTRS offsets are zero-based record positions: the next page above starts at 25, not 1. The currently supported GET search interface is marked deprecated in NASA’s OpenAPI schema; the adapter’s endpoint should be rechecked if NASA removes that interface. No entire report corpus is crawled automatically.

When proposing catalogue fields, distinguish measured output, rated capacity, design requirements and simulations. Record beginning/end-of-life state, temperature, distance from the Sun, array versus bus/payload power, units and the exact supporting passage. An abstract is a discovery lead, not automatic authorization to fill a satellite’s power rating. Follow Satellite and space-power catalogue for evidence review and conservative merging.

Complete satellite metadata with DISCOS#

Create a free ESA Space Debris User account, accept the DISCOSweb terms and obtain an API token. The library never creates an account or reads environment variables implicitly. Pass the token explicitly:

import os
import orbitalpv as opv

with opv.DiscosConnector(token=os.environ["DISCOS_TOKEN"]) as api:
    matches = api.by_norad(25544, cache_path="data/research/discos-iss.json")
    print(matches.data["data"])
    # For a known DISCOS internal ID, use api.object(discos_id).

NORAD numbers and DISCOS internal IDs must not be interchanged. Empty and multiple matches remain visible. Token headers are not written to receipts. Respect ESA’s account terms before redistributing metadata.

Discover and download published research datasets#

import orbitalpv as opv

with opv.ZenodoConnector() as api:
    found = api.search("OPSSAT", page=1, page_size=10)
    dataset = api.record(15108715)  # Fixed OPS-SAT-AD release
    print(dataset.data["metadata"]["license"])
    api.download_file(15108715, "requirements.txt", "data/research/requirements.txt")

with opv.DataverseConnector() as api:
    found = api.search("Flying Laptop", start=0, page_size=10)
    description = api.dataset("doi:10.18419/darus-456", version="1.0")
    api.download_file(
        "doi:10.18419/darus-456", "FLP_TM.pdf", "data/research/FLP_TM.pdf",
        version="1.0",
    )

Zenodo pages start at 1 and anonymous page size is capped at 25. Dataverse’s start is a zero-based record offset; supply base_url for another Dataverse installation. Pin published numeric versions for reproducibility. Downloads verify provider checksums, preserve license/version context in sidecars, and default to a 100 MB limit. Inspect metadata before deliberately raising max_bytes. Dataverse requests original files so tabular conversion does not silently invalidate the provider checksum. Files are never executed or extracted. These generic adapters complement the specialized Flying Laptop and OPS-SAT loaders, which already know their measurement formats.

Retrieve geometry and mission products#

import orbitalpv as opv

with opv.HorizonsConnector() as api:
    ephemeris = api.vectors(
        "-61", start="2024-01-01", stop="2024-01-04", step="1 d",
        cache_path="data/research/juno-vectors.json",
    )
    print(ephemeris.data["result"])

with opv.PdsConnector() as api:
    bundles = api.search('(product_class eq "Product_Bundle")', limit=10)
    if bundles.data["data"]:
        product = api.product(bundles.data["data"][0]["id"])

with opv.PsaConnector() as api:
    products = api.query(
        "SELECT TOP 10 granule_uid, instrument_host_name, access_url "
        "FROM psa.epn_core WHERE instrument_host_name = 'Mars Express'",
        max_records=10,
    )
    print(products.data["metadata"], products.data["data"])

Horizons defaults are Sun-centered geometric vectors, ICRF axes, AU/day units and TDB dates; they must not be relabeled as UTC. The returned text contains the provider’s original headers and ephemeris. Ambiguous targets or unavailable coverage raise an error. These vectors do not supply array attitude or shading.

PDS uses its own q expression syntax. For cursor pagination, pass both sort and search_after. Its documented keywords parameter returned unfiltered records during live verification, so this connector deliberately exposes explicit expressions instead. Empty results can mean incomplete archive coverage. PSA is a bounded SELECT interface; a result reaching max_records may be truncated. For both archives, inspect product labels, calibration, units and access terms before downloading or interpreting data.

Add space-weather event context#

import orbitalpv as opv

with opv.DonkiConnector() as api:
    events = api.events(
        "FLR", start="2024-05-10", stop="2024-05-11",
        cache_path="data/research/flares.json",
    )
    for event in events.data or []:
        print(event["beginTime"], event["classType"], event["link"])

Supported event types are CME, FLR, SEP, GST, IPS, MPC, RBE and HSS. UTC date endpoints are inclusive under DONKI’s convention. The raw empty response may be null or a list. This uses CCMC’s direct free endpoint and does not need an api.nasa.gov key. The base host follows the September 2026 migration to ccmc.gsfc.nasa.gov. Follow CCMC’s research-use and attribution policy.

Reproduce the access check#

The Free spacecraft research APIs: evidence, archives and geometry example runs offline against small original responses captured from all seven anonymous APIs. From an installed checkout, explicitly refresh real queries with:

python scripts/check_research_connectors.py --cache-dir data/research-connectors
# Optional, after configuring a free DISCOS account/token:
python scripts/check_research_connectors.py --cache-dir data/research-connectors --discos

The script writes a machine-readable access report and exits nonzero if any selected provider fails. These new connector examples use committed snapshots during CI and documentation builds.

Primary interface references#