Once a missile is in free flight, every second of additional tracking data halves the uncertainty ellipse around its predicted impact point. Ground-based radars cover limited arcs, can be jammed or blinded, and hand off late in the threat timeline. A satellite constellation that continuously observes the mid-course and terminal phases from multiple viewing angles feeds a persistent, globally coherent picture that no ground network can replicate at the same latency or coverage breadth.
The satellite stack combines infrared sensors for plume tracking in boost and early mid-course, with RF signal collection and precise optical astrometry for mid-course and re-entry. On-board Kalman filter and trajectory propagation algorithms reduce the raw observation stream to a predicted impact centroid and a confidence ellipse before the data even hits the ground. That on-orbit compute step is decisive: it cuts the round-trip latency from sensor to decision-maker from minutes to seconds.
The operational outcome is a decision window that actually exists. Civil defence controllers receive geo-referenced impact probability maps in time to order shelter-in-place or evacuation of a specific district rather than an entire city. Interceptor batteries receive refined aim-point cues that improve single-shot kill probability. The intelligence product persists after the event as a forensic trajectory reconstruction that anchors diplomatic and legal responses.
Frequently asked
What is the difference between trajectory prediction and simple launch detection?
Launch detection identifies the ignition event — a bright infrared flash — within seconds of motor ignition. Trajectory prediction is everything that follows: computing the missile's current velocity vector, fitting a propagated arc through atmospheric and gravitational models, and outputting a probabilistic impact ellipse. Detection tells you something was fired; trajectory prediction tells you where it is going and when. The two functions are complementary but architecturally distinct.
Can a small nation afford a sovereign trajectory-prediction satellite layer?
Not in isolation at the same sensor quality as SBIRS or NGG OPIR, which exceed $2 billion per satellite. However, a proliferated LEO constellation of microsatellites carrying medium-wave infrared (MWIR) payloads can be built for $80–200 million per satellite at present commercial rates, and a 6–10 satellite constellation targeting regional threats is financially achievable for mid-tier economies. The trade-off is narrower spectral range and lower sensitivity, partially offset by shorter slant ranges in LEO.
Why can't we just buy trajectory data as a service from a US or allied provider?
Several vendors and government agencies offer threat data sharing through frameworks like the Combined Space Operations Centre (CSpOC), but that data is released on terms controlled by the originating nation, filtered for classification, and can be withheld or delayed during periods of geopolitical tension. More critically, if a missile is inbound, decisions about intercept authorisation and population alerting are made in seconds — a service agreement with a foreign entity cannot substitute for real-time sovereign sensor custody and a sovereign command chain.
What orbits work best for trajectory prediction payloads?
GEO provides persistent staring geometry ideal for the boost phase through early mid-course, but its 35,786 km altitude reduces sensitivity to smaller, cooler missiles. LEO constellations at 1,000–1,200 km provide roughly 100× greater radiance at the focal plane compared to GEO, enabling detection and tracking of non-emitting glide vehicles and cooler solid-fuel motors. The current US architecture layers both: GEO for wide-area cuing, proliferated LEO for precision tracking and discrimination — a model sovereign designers should study.
How does a trajectory prediction system handle hypersonic glide vehicles, which don't follow a ballistic arc?
Classical Keplerian propagators fail for hypersonic glide vehicles (HGVs) because they manoeuvre laterally and depressively in the upper atmosphere. Modern prediction engines use adaptive Kalman filters or particle filters that do not assume a fixed ballistic coefficient, continuously re-fitting the vehicle's aerodynamic state from successive sensor measurements. The challenge is that HGV tracking demands more frequent measurement updates — ideally every 1–3 seconds — placing higher demands on constellation revisit and cross-link data rates. This is addressed in the companion page on Hypersonic Glide Tracking.
What ground infrastructure is required to process tracking data into actionable impact predictions?
You need at minimum: a hardened mission ground station with sub-second contact windows to the constellation, a classified mission processing enclave running a multi-hypothesis track-before-detect algorithm, a connection to a national atmospheric data feed (e.g. from a WMO-affiliated NWP centre), and a certified interface to the national missile defence command authority. Many nations underestimate the ground segment cost, which can rival the space segment at 40–60% of total programme cost.
What level of tracking accuracy is needed for a useful intercept cue?
The Missile Defense Agency's performance benchmarks suggest that an intercept cue must place the predicted impact point within a 1-sigma ellipse of roughly 150–300 m to support high-confidence THAAD or Patriot engagement planning. Accuracy degrades with range and latency; a nation with a single continental ground station receiving data at 2-second latency might see real-world performance at 500–800 m 1-sigma, which still supports area-defence intercept but not point-defence of a hardened facility.
How is this capability governed internationally — are there treaty obligations?
Trajectory prediction satellites are military space systems and sit outside the governance of bodies like IMO or ICAO. Their operation is bounded by the Outer Space Treaty (1967), which prohibits weapons of mass destruction in orbit but places no restriction on early-warning constellations. ITU-R frequency coordination applies to the radio links, and nations must file orbital slots and frequency assignments with the ITU. There is no multilateral treaty specifically governing missile warning data sharing, though bilateral agreements (e.g. US–Japan, US–Australia) operationalise cooperative cueing.