20 independent data streams across electromagnetic, atmospheric, infrastructure, ecological, and geophysical domains. Zero editorialization. Every observation timestamped, source-attributed, hash-verified, and anchored to the Trust Layer. We don't tell you what to think β we show you what the instruments are reading.
If you just got here, here's what you need to know. This isn't a weather app. It's a multi-domain observation platform. Think of it as 20 different scientific instruments all feeding into one dashboard β and the dashboard is watching for moments when those instruments say the same thing at the same time.
The bottom line: this platform doesn't tell you what to think. It shows you what the instruments are reading, flags when multiple instruments agree something unusual is happening, and then reminds you that coincidence is statistically expected. Everything else is up to you.
Truth Observatory is a real-time multi-domain observation platform that collects data from 20 public government and scientific data sources spanning five domains:
Electromagnetic β solar flux, geomagnetic indices, ionospheric TEC, Schumann resonance, ionospheric heater schedules.
Atmospheric β NEXRAD radar, GOES satellite imagery, surface stations, deposition chemistry, precipitation.
Infrastructure β power grid demand (EIA-930), cell tower registries (FCC ASR), aircraft tracking (ADS-B), NOTAMs.
Ecological β pollinator observation indices, trace metals in waterways, atmospheric deposition monitoring, active wildfire incidents (NIFC).
Geophysical β USGS seismic activity, USGS volcanic activity monitoring (alert levels, aviation color codes, threat assessments).
The platform models statistical baselines for each stream, detects deviations from normal, and cross-correlates anomalies across all domains. Solar flare β geomagnetic spike β ionospheric disturbance β grid demand anomaly β four domains, one observable chain. That's the 10-mile-high view, built from the bottom up.
Every piece of data is pulled directly from the source β NOAA, USGS, FAA, EPA, EIA, SWPC, and independent scientific networks. No data is generated, interpolated, or editorialized. If a feed goes offline, it's recorded as a gap, not silently skipped.
The platform runs on a single Hetzner VPS in Finland for approximately $8/month. All data sources are free, public APIs. The entire system is designed to be reproducible β same input data always produces the same output.
The Observatory pipeline is deterministic. Every stage is auditable and every output is traceable to its source manifest.
Every feed pulls from a public, freely accessible source. No API keys, no paywalls, no proprietary data.
AtmosCore is the composite state engine that distills all 20 feeds into a single observation suitability score. It computes four physical primitives, each derived from a subset of feeds:
The 4/42 designation refers to the four physical primitives grounded in the 42-Doctrine Deterministic Dissolution Ladder β the architectural framework that guarantees deterministic outputs across the entire system.
No interpolation. If data is missing, it's recorded as a gap. We never fill holes with estimates.
No editorialization. The system shows you what the instruments read. It does not tell you what it means.
No AI judgment. Baselines are statistical. Deviations are mathematical. Correlations are temporal. None of this requires or uses machine learning.
Full provenance. Every observation is traceable from the UI β API β Event Ledger β manifest β source URL β raw file with SHA-256 hash.
On-chain verification. Ledger block hashes are anchored to the Trust Layer (DWTL) β an independent Layer 1 blockchain. Once anchored, observations cannot be retroactively modified, deleted, or reinterpreted. The blockchain is not the product β it is the proof layer.
Observatory generates deterministic SHA-256 hashes for every observation. Every 15-minute block's Merkle root is submitted to DWTL β the ecosystem's Layer 1 blockchain at dwtl.io.
This anchoring provides four guarantees:
1. Tamper-proof observations. Once a block hash is on-chain, the raw data that produced it cannot be retroactively altered without breaking the Merkle proof.
2. Safe pruning. Raw observation files are pruned at 30 days, but the on-chain proof of their existence is permanent. Anyone can verify an observation existed by checking its hash against the block's Merkle root.
3. Immutable verdicts. The skeptic engine's correlation verdicts are recorded with the block hash. A "COINCIDENCE" verdict cannot be retroactively changed to "CONFIRMED" β the original judgment is on-chain.
4. Independent verification. Any third party can reconstruct the Merkle tree from raw manifests and verify it matches the on-chain root. No trust in DarkWave Studios is required β the math is the proof.
Observatory is designed to grow. If you operate a sensor network, maintain a public data feed, or have access to a scientific data source we don't yet cover β you can propose it for inclusion.
The contribution model is straightforward:
1. Propose a feed. Describe your data source, its update frequency, what it measures, and how to access it (must be freely accessible β no API keys, no paywalls).
2. Define a manifest schema. Specify the fields your collector will write to its NDJSON manifest. At minimum: type, source, retrievedAt, sha256.
3. We validate and integrate. Your collector is added to the supervisor fleet. Your data enters the baseline engine, the deviation detector, the cross-correlator, and the Event Ledger. You get full access to the correlation output across all feeds, including yours.
The more feeds, the richer the correlation model. Independent sensor networks β air quality, electromagnetic field measurements, acoustic monitoring, water quality β all contribute to a more complete picture of what the atmosphere is actually doing.
All Observatory data is freely accessible via REST API. No authentication required.
Observatory currently runs on a $8/month Hetzner VPS. The entire platform β 20 collectors, the correlation engine, the API server, and the dashboard β operates within this budget using exclusively free public data sources.
What funding enables:
More feeds. Weather balloon radiosonde data, ocean buoy networks (NDBC), volcanic gas emissions (USGS), aurora cameras, magnetometer networks, and space weather beyond SWPC (DSCOVR, ACE real-time solar wind).
Longer retention. Currently raw data is pruned at 30 days. With additional storage, we can retain 1 year+ of raw observations for long-term trend analysis and seasonal correlation patterns.
Real-time alerting. SMS/email/webhook notifications when deviations exceed configurable thresholds or when new cross-feed correlations are detected.
Regional coverage. Per-zip-code local condition dashboards, station-level data overlays, and location-aware deviation alerts.
Mobile app. Native iOS/Android application for field researchers and independent observers.
Raw observation files are pruned after 30 days to manage storage on the current infrastructure.
Manifest metadata β including timestamps, source attribution, SHA-256 hashes, and observation summaries β is retained permanently. This means every observation that ever passed through the system can be verified by its hash, even after the raw file is pruned.
On-chain anchoring β 15-minute block Merkle roots are submitted to DWTL. Even if this server is destroyed, the cryptographic proof of every observation's existence is permanent on Layer 1.
Event Ledger blocks and their hashes are permanent. The full ledger history is accessible via /api/ledger and searchable via the Event Explorer.
Correlation patterns are recomputed every 5 minutes from the rolling 7-day window. Historical patterns are preserved in the Event Ledger.