Indonesian SBAS ION Paper Iterations

Scope status

This note tracks the iterative research paper series that reframed initial IRI-2020 model validation into ION-style SBAS ionospheric threat modeling. It is a meta-research note, not a source note.

Boundary:

  • It does not restate the full paper content (that remains in the LaTeX sources).
  • It preserves the evolution of research framing across iterations.
  • It links empirical results to SBAS integrity implications.

Iteration timeline

VersionDateFraming shiftKey artifact
V0Apr 2026IRI-2020 validation onlyionosphere_model_error_indonesia.tex
V1Apr 2026SBAS threat discovery frameworkion_indonesia_sbas_ro_threat_paper.tex
V2Apr 2026Expanded dataset (5→8 days), QC, coverage gapsion_indonesia_sbas_ro_threat_paper_v2.tex
V3Apr 2026Space-weather + magnetic coordinates + threat budgetion_indonesia_sbas_ro_threat_paper_v3.tex
V4Apr 2026IGRF14 magnetic-field upgrade, best model: R²=0.44ion_indonesia_sbas_ro_threat_paper_v4.tex

V1 → methodology correction

The first ION-style paper avoided overclaiming that Spire podTec mean values were certified SBAS vertical ionospheric delays. It consistently labels the observable as an RO-derived occultation TEC proxy.

V2 → dataset expansion and coverage diagnosis

  • Expanded from 5 to 8 stratified 2025 days
  • Added explicit quality-control accounting
  • Added local-time coverage-gap quantification
  • Key finding: 18–21 LT post-sunset risk-window strict profiles = 0

V3 → space-weather and magnetic-coordinate upgrade

  • Added GFZ Kp, Kyoto Dst, F10.7
  • Added centered-dipole magnetic latitude + magnetic local time
  • Replaced sparse geographic model with additive robust Huber regression
  • Added pre-operational SBAS threat-budget scaffold

V4 → IGRF14 magnetic upgrade

  • Replaced centered-dipole with ppigrf-derived IGRF14 dip latitude
  • Added full IGRF field diagnostics
  • Best model: RMSE 36.2 TECU, R² 0.44, p99 residual 114.2 TECU (18.5 m L1)

Persistent integrity finding across all versions

The post-sunset 18–21 LT coverage gap remains 0 profiles in every iteration. This is treated as:

  • An acquisition priority for future RO campaigns
  • An integrity penalty that must be accounted for in any threat budget
  • Not evidence of low risk — absence of data is not evidence of absence

As defined in the V4 paper:

  1. Targeted data expansion: acquire days specifically designed to fill 18–21 LT post-sunset coverage
  2. Add storm-event days using Kp/Dst-driven selection
  3. Build an acquisition diagnostic that selects days by local-time coverage and geomagnetic disturbance before downloading
  4. Re-run V4 IGRF pipeline on expanded targeted sample
  5. Only then begin candidate GIVE-like regional bin design

See also