Source - GNSS-RO Indonesia Empirical Study
Scope of this note
This is a starter source scaffold for the Indonesian GNSS Radio Occultation empirical study conducted in April 2026.
Boundary:
- It is not a completed source extraction.
- It does not independently verify the numerical results; it documents what the vault contains.
- It exists to anchor empirical claims in the vault to a recognizable study context.
Why this source note matters
The vault contains a series of research artifacts (LaTeX papers, analysis scripts, figures, CSV data) produced during a systematic GNSS-RO-based study of Indonesian ionospheric conditions. Without a study-level source scaffold, it is unclear which claims are derived from this dataset versus other sources.
Study provenance
| Field | Value |
|---|---|
| Conducted | April 2026 |
| Data source | Spire Global constellation via UCAR CDAAC |
| Period | 5 representative days (Jan–May 2025); later expanded to 8 stratified days |
| Region | Indonesia geofence: −11.0° to 6.0° N, 95.0° to 141.0° E |
| Software | Custom Python scripts |
| Paper series | ION-style SBAS ionospheric threat papers (V1–V4) |
Key numerical findings
Initial 5-day study
- 301 Indonesia occultations from 17,909 source files
- IRI-2020 bias: 66.72 TECU (87.6% mean underestimate)
- RMSE: 94.34 TECU; R²: −0.73
Expanded 8-day study (V2)
- 463 intersecting profiles → 208 strict tangent-point profiles → 207 QC-valid profiles
- RO TEC p95: 166.0 TECU; p99: 341.1 TECU
- Leave-day-out RMSE: 56.6 TECU; R²: −0.38
Space-weather upgrade (V3)
- Added GFZ Kp, Kyoto Dst, F10.7, centered-dipole magnetic coordinates
- Best model R²: 0.42; p99 residual: 120.1 TECU (19.5 m L1)
IGRF upgrade (V4)
- Replaced centered-dipole with IGRF14 dip-latitude
- Best model R²: 0.44; p99 residual: 114.2 TECU (18.5 m L1)
Post-sunset coverage gap
- 18–21 LT strict profiles: 0 in all iterations
- Flagged as integrity-relevant limitation, not evidence of low risk
Current in-vault references supported by this scaffold
- IRI-2020 vs GNSS-RO Indonesia — direct comparison results
- SBAS Ionospheric Threat — Empirical Evidence — threat quantification
- GNSS Radio Occultation — technique application
- Total Electron Content (TEC) — observable statistics
- Indonesian SBAS ION Paper Iterations — paper series tracking
- Indonesia — country-specific empirical anchor
Claims this scaffold does NOT yet support
- Any claim that the findings generalize beyond the sampled days
- Any claim that the sampling is sufficient for operational GIVE design
- Any claim that Spire podTec accuracy is certified for SBAS use
- Any ranking of Indonesia against other equatorial regions
Immediate audit questions
- Which specific Spire mission(s) and data product version(s) were used?
- Were formal UCAR CDAAC quality-control flags applied uniformly?
- How was the IRI-2020 model invoked (library, API, online tool)?
- What preprocessing was applied to podTec before TEC extraction?
- Which LaTeX/paper version is the primary citation target?
Suggested downstream cleanup targets
If this scaffold becomes source-backed, revisit:
- IRI-2020 vs GNSS-RO Indonesia
- SBAS Ionospheric Threat — Empirical Evidence
- Indonesia — empirical claims in the country note
Extraction template for next cycle
- Exact dataset URLs or DOIs
- Software version and dependency list
- Sample-size justification
- Statistical methodology references
- External validation or peer-review status