Defense & Collection Planning · Open Source

Defense Satellite Intelligence and Collection Planning From Open Sources

Unclassified GEOINT and OSINT for defense and security analysts and collection managers: where imaging is worth it, the SAR and optical plan for an event, when the place can next be seen, and a sourced brief — decision support, not targeting.

Open the Watchfloor Predict satellite passes

Quick answer — what Off-Nadir Delta is

Open-source collection planning decides which place needs imagery, from which class of sensor, and when that sensor can next look — from open reporting and open satellite catalogs rather than classified collection. Off-Nadir Delta classes each event by the resolution it needs, builds the SAR and optical plan for one event, predicts passes for 13 free and commercial satellite families, checks SAR terrain geometry, and returns sourced, caveated briefs. Open sources only; not for targeting or the surveillance of individuals.

The scope, stated plainly

What this is for

Deciding where open or commercial imagery is worth requesting
Situational awareness of conflicts and crises from open sources
Verifying open-source claims with imagery
Watching facilities and areas of interest over time

What it is not

Not a classified system — open sources only
Not an operational targeting tool
Not for the surveillance of individuals
Not a tasking service — access windows are opportunities, not captures

Imaging priority

Each event is classed by the resolution it needs — coarse, free Sentinel-class, or commercial sub-metre — and shown as a heat layer on the Watchfloor map. Events that cannot be placed as a point, or that imagery cannot resolve, are left out.

Event imagery plan

SAR and optical, each searched once against the event footprint, with pre/post bracketing and the SAR change-pair status. If an event is located only to a country or province, no plan is built.

Next look

Pass prediction for 13 families, free systematic versus taskable commercial. Geometric access only — never presented as a confirmed acquisition.

Terrain

SAR layover and shadow share for a given incidence and look direction, and line of sight between two points (REST API and Python SDK). Elevation hillshade on the map.

Sourced briefs

Delta Agent answers in plain language with sources and a sensor recommendation, and declines to assess when it finds no evidence.

Facility-scale watch

Watch a base or airfield from the facility registry, or any point: measured by default from Sentinel-1 radar backscatter, with anomalies against the site’s own history.

How It Works

Decide where imaging is worth it

Each event is classed by the resolution it needs: coarse (≤100 m), Sentinel-class (≤10 m, free), or sub-metre (commercial). The Watchfloor map shows this as an imaging-priority heat layer, and on plans that include it the list sorts by Imaging value.

Get the SAR and optical plan

For one located event, radar and optical are each searched once against the event footprint, with pre- and post-event scenes bracketed and the SAR change-pair status stated. A sub-metre recommendation never cancels what the free catalog already shows.

Know when it can next be seen

Predict passes for 13 satellite families, free systematic collectors separated from taskable commercial platforms. Check SAR layover and shadow, or line of sight between two points, from the terrain model.

Brief, then keep watch

Ask Delta Agent for a sourced, caveated brief. Then watch the base, airfield or area in Delta Watchlist: measured from new imagery, with an alert when a new scene is published.

What the event record tells you before you collect

How precisely it is located

Every event sits on a ladder: reported (country or province), localized (town) or pinpointed (block or facility), with the reason it stopped where it did. Coordinates that contradict the reported country are dropped rather than plotted.

How much weight the reporting carries

The best report on each event carries a NATO (Admiralty) grade — source reliability A–F and information credibility 1–6 — kept separate from how many outlets carried it. Breadth is not independence, and neither is stated as such.

Building a pipeline? The REST API and Python SDK cover the collection reads end to end — see the developer docs.

Frequently Asked Questions

What is collection planning from open sources?▼

It is deciding which place needs imagery, from which kind of sensor, and when that sensor can next look — using open reporting and open satellite catalogs instead of classified collection. Off-Nadir Delta answers each part: which events need coarse, Sentinel-class or sub-metre imagery, which SAR and optical scenes already exist before and after an event, and which satellites have access in the coming days.

Which satellites does pass prediction cover?▼

Thirteen families. Free systematic collectors — Sentinel-1, Sentinel-2, Landsat and NISAR — are kept separate from taskable commercial platforms: WorldView, ICEYE, Capella, SkySat, Umbra, Synspective, iQPS, RADARSAT-2 and COSMO-SkyMed. The prediction uses public orbital elements within a 7-day window, so it gives geometric access only: a commercial access window is an opportunity to task, not a scheduled capture. If orbital data cannot be retrieved, the answer says timing is unavailable instead of reporting no passes.

Can free imagery count vehicles at a base?▼

No. Sentinel-1 and Sentinel-2 work at roughly 10 m, which supports facility-scale activity — apron occupancy, large concentrations, change over time — but not counting or typing individual vehicles; that needs sub-metre commercial imagery. Delta Agent says this rather than inventing a count, and the imagery plan states when sub-metre collection is recommended and when the question cannot be decided from the event record.

How does terrain affect a SAR collection?▼

Steep ground facing the radar folds over (layover), and ground facing away is hidden (shadow). For an area, an incidence angle and a look direction, the terrain analysis returns the share of the area lost to each, from the Copernicus 30 m elevation model. It also returns a line-of-sight verdict between two points, including Earth curvature. Both are available in the REST API and the Python SDK; the map shows the same elevation model as a hillshade layer.

Is this classified intelligence or a targeting tool?▼

No. Off-Nadir Delta uses only publicly available sources and is decision support for situational awareness, monitoring and the verification of open-source claims. It is not a classified system and is not designed for operational targeting or for the surveillance of individuals. Where open imagery or open reporting cannot support a judgment, the product says so.

Can I run this from my own pipeline?▼

Yes. The REST API and the Python SDK cover the same collection reads — imaging priority, the event imagery plan, pass prediction and terrain — and the MCP server gives AI agents imaging priority, the event plan and pass prediction. See the developer docs.