Search BackCountrySignal

    Open full search pageSearches run in your browser. Nothing you type is sent anywhere.

    COMMUNITY DATA 2.0

    Build on
    the evidence.

    Use reviewed public field records and intentionally public node data in your own analysis. Private reports, contact details, review notes and exact private sites are excluded.

    GET endpointResponseEvidence standard
    /api/community.jsonversion, generated_at, reports[]All approved public reports; structured evidence included
    /api/field-evidence.csvCSV downloadSame approved public dataset; spreadsheet-safe text cells
    /api/nodesnodes[]Curated public records; status must be checked; hidden coordinates removed, area coordinates rounded to 2 decimals
    /api/v1/stats/overviewobservations, hardware, rolesPublic MQTT feed statistics; observations are not verified deployments

    Read the dataset

    const response = await fetch('https://backcountrysignal.org/api/community.json');
    if (!response.ok) throw new Error('Evidence unavailable');
    const data = await response.json();
    const successfulTests = data.reports.filter(
      report => report.report_type === 'field-test' && report.outcome === 'success'
    );

    Report fields

    id, node_id, report_type, area, observed_at, result, radio, antenna, terrain, distance_miles, created_at, outcome, evidence. Outcome is success, failure, partial or unknown. Older evidence remains unknown until reviewed; don't infer success from distance alone.

    Structured evidence

    link_id, hardware_id, modem_preset, endpoint_a_area, endpoint_b_area, antenna_a_ft, antenna_b_ft, elevation_a_ft, elevation_b_ft, packets_sent, packets_received, rssi_dbm, snr_db, planned_assessment. Missing values are null. Elevations are feet MSL; antenna heights are feet AGL. Link ID associates a test with a locally saved plan, without publishing its exact coordinates.

    Use responsibly

    Public GET endpoints support cross-origin reads and return JSON with no-store. Check HTTP status; failures return an error instead of a zero count. Take periodic snapshots rather than polling rapidly. Cite BackCountrySignal and the record IDs in derived work. No broader data license is implied for third-party MQTT observations. Public narratives are reviewed for private details before approval.