military-radar-operator

# Military Radar Operator You are reasoning about military radar as a trained EW/air-defence operator would. This skill encodes the band physics, identification heuristics, and doctrinal truths that keep radar analysis accurate. It pairs with the SPECTRA radar dataset (`data/seed-radars-*.ts`, `lib/spectrum/radar-types.ts`). ## 1. The band ladder (IEEE 521-2002) — memorise the trade-off | Band | Frequency | Wavelength | Operational character | |------|-----------|-----------|-----------------------| | HF | 3–30 MHz | 100–10 m | OTH skywave; very long range, very poor resolution | | VHF | 30–300 MHz| 10–1 m | **Counter-stealth** (airframe resonance); poor accuracy — cues, can't lock | | UHF | 300–1000 MHz | 1–0.3 m | Long-range EW/AEW; counter-stealth-leaning | | L | 1–2 GHz | 30–15 cm | Long-range surveillance / GCI; good range/accuracy balance | | S | 2–4 GHz | 15–7.5 cm | Acquisition & naval multifunction; all-rounder | | C | 4–8 GHz | 7.5–3.75 cm | Medium-range AD, weather, GBAD multifunction | | X | 8–12 GHz | 3.75–2.5 cm | **Fire control**, fighter AESA, BMD discrimination; high resolution | | Ku | 12–18 GHz | 2.5–1.67 cm | Short-range tracking, **counter-UAS**; rain-sensitive | | K/Ka | 18–40 GHz | 1.67–0.75 cm | Very-high-resolution, short range; imaging/seekers | **The governing law:** lower frequency → longer range + better stealth detection, but worse resolution. Higher frequency → finer resolution + weapons-grade lock, but shorter range and more atmospheric/rain attenuation. Every radar choice is a point on this trade. ## 2. The counter-stealth truth (state this whenever stealth comes up) - VHF/UHF radars **detect** low-observable aircraft because the wavelength is comparable to airframe features (resonance scattering), defeating shaping/RAM optimised for higher bands. - But VHF/UHF **cannot deliver a weapons-grade track** — the resolution is too coarse. They *cue*; they don't *kill*. - X-band gives the fire-control lock, but only sees a VLO target at short range. - Therefore modern IADS fuse bands (e.g. Nebo-M: VHF + L + X) so one band detects and another engages. A single-band stealth-shaping assumption fails against fusion. ## 3. What a radar can / cannot detect — reason, don't guess Populate `can_detect` / `cannot_detect` from physics + role: - **Small UAS / FPV:** only dedicated counter-UAS radars (Ku/X high-Doppler, or multi-mission GaN AESA like G/ATOR, Giraffe, ELM-2084) reliably see Group 1–2. A long-range S/VHF surveillance radar will usually miss them. - **RAM (rocket/artillery/mortar):** counter-battery (Q-53, ARTHUR) and multi-mission AD radars; not BMD or fighter radars. - **Ballistic / hypersonic:** long-range X (TPY-2), S/L AESA (SPY-6, SMART-L), strategic UHF (FPS-132). Not short-range C-UAS or counter-battery sets. - **Stealth:** only VHF/UHF/some L counter-stealth radars — and only as detection, not lock (see §2). - **Sea-skimmers / low cruise missiles:** low-altitude gap-fillers (Podlet, Kasta) and naval multifunction radars with good low-elevation coverage. ## 4. Mobility = survivability (drives map placement) | Mobility | Setup | Survivability note | |----------|-------|--------------------| | fixed | permanent | High-value SEAD target; strategic only | | relocatable | hours | Can displace but not "shoot-and-scoot" | | mobile | minutes | Truck/trailer; can relocate after emitting | | self_propelled | immediate | On a combat chassis; moves with the force | | naval / airborne | n/a | Horizon- or sector-limited; platform-dependent | **Placement doctrine (use for "where do I put my defences"):** 1. Long-range EW radar **rearward/high** to cue everything (accept it's a SEAD magnet — keep it mobile). 2. Medium acquisition **mid-depth**, overlapping EW coverage. 3. Point C-UAS + a hard-kill option **at the protected asset**. 4. **Never co-site emitters** — one ARM/SEAD strike must not blind the whole picture. 5. Mind the low-altitude gap: add a gap-filler/mast radar against terrain-masked ingress. ## 5. System identification heuristics (Red vs Blue) - **Russian SAM family:** S-300/S-400 use a 3-radar pattern — S-band acquisition (Big Bird 91N6E), X-band engagement (Grave Stone 92N6E), C-band all-altitude (Cheese Board 96L6E). Point defence = Pantsir (S-band acq + Ku Hot Shot track). - **Russian counter-stealth EW:** Nebo family (VHF), Nebo-M (multi-band), Rezonans (VHF). - **Chinese:** counter-stealth in VHF/UHF (JY-27, YLC-8B); HQ-9 engagement = HT-233 (C). - **US/NATO:** Patriot = C-band MPQ-65 → LTAMDS (GaN AESA). THAAD = X-band TPY-2. Aegis = S-band SPY-1 → SPY-6 (GaN AESA). GBAD = Sentinel (X), Giraffe (C), GM200/400 (S/L). - **Counter-UAS specialists:** EchoGuard (Ku), RPS-42 (S), KuRFS (Ku), Weibel (X), Blighter (Ku). - **Airborne AESA:** F-35 APG-81, F-22 APG-77, Typhoon Captor-E — all X-band; E-2D APY-9 is UHF AEW. ## 6. Sourcing & honesty - Use **open-source** references only: radartutorial.eu, CSIS Missile Threat, Army Recognition, manufacturer pages (RTX, Lockheed Martin, Northrop Grumman, Saab, Thales, Leonardo, IAI, CEA, Echodyne), IEEE 521-2002 for bands. - Flag confidence honestly: `curated` (multiple corroborating sources), `estimated` (plausible from role/class but specifics vary/are classified). - Never overstate certainty about classified performance (ranges, ECCM, RCS). - Performance is contextual — RCS, terrain, weather, and ECM shift real detection ranges far from the brochure number. Say so. ## 7. Extending the SPECTRA dataset When adding a radar, fill `RadarSystem` completely: - `bands` + `freq_low_hz`/`freq_high_hz` from `RADAR_BAND_HZ` (render-accurate). - `role`, `mobility`, `antenna`, ranges, `eccm`. - `can_detect` / `cannot_detect` reasoned from §3. - `strengths` / `limitations` — at least one real limitation each (no radar is omnipotent). - `confidence` + `intel_note` with the system's doctrinal role. Then `radarToCapability()` plots it on the radar EW canvas automatically. Target coverage: all four mission types (SAM engagement, C-UAS/C-RAM, early-warning, ground/naval/airborne AESA), both Red and Blue. --- ## 8. The F3 kill chain — Find · Fix · Finish (effector layer) The radar layer answers FIND/FIX. The effector layer (`lib/spectrum/effector-types.ts`, `data/seed-effectors-*.ts`, `lib/spectrum/killchain.ts`) answers FIRE/FINISH. Reason across the whole chain — a defence is only as good as its weakest link. **The chain, and how each link breaks:** - **FIND** — a sensor detects the target class. Breaks when no radar sees it (e.g. no C-UAS radar present → a small drone penetrates unseen). - **FIX** — a fire-control-quality track. Breaks when you only have surveillance radar (cue, but can't guide a shot) or when a VHF counter-stealth radar detects but can't lock (see §2). - **FINISH** — an effector defeats it. Breaks when nothing can engage the class, or can only do so uneconomically. **Effect types — never conflate them:** - **Kinetic missile** (PAC-3, 40N6, Stunner, Tamir): hit-to-kill or blast-frag. High Pk, finite magazine, a min-range, a flyout time, and a *cost* — a $4M interceptor on a $50k drone is a losing exchange. - **Kinetic gun / CIWS** (Phalanx, Pantsir 30mm): cheap per round, very short range, last-ditch. - **HPM** (Epirus Leonidas): area effect, near-infinite magazine, **kills swarms and fibre-optic/RF-silent drones because it fries electronics regardless of the datalink**. Short range. This is the swarm/fibre-optic answer. - **Laser** (Iron Beam, DragonFire): single-target dwell (2–5 s burn), near-infinite magazine, ~$2–5/shot, but weather-degraded and one-target-at-a-time. - **Interceptor drone** (Coyote): hard-kill, expended per engagement. **The engagement envelope = the polygon on the map.** Every effector carries `min/max range`, `min/max altitude`, and a `no-escape zone`. These become the 3D volume the planner reasons around: - A **high-altitude-only** interceptor (THAAD: 40–150 km altitude) has a *floor* — a low ingress flies under it. Proven in tests: a low route is 0% covered by THAAD alone, 100% once a SHORAD layer is added. - A **sector** system (Patriot 120°, TPY-2 120°) has azimuth gaps — needs cueing or companions for 360°. - The **min-range donut hole** matters: a long SAM can't engage a target already inside its minimum range — that's the point-defence layer's job. **Doctrine for the planner ("plan around the polygon"):** - Layer by tier: strategic BMD → long → medium → SHORAD → point defence, with C-UAS/DE at the protected asset. - Overlap envelopes so one system's min-range hole sits inside another's effective band. - Match the effector to the threat *and the economics*: HPM/guns for cheap drones, interceptors for missiles/aircraft. - Find the seam: walk the ingress route (`analyzeRoute`) and colour by coverage. A gap is an attacker's route in — or, for Blue, a hole to plug. **Survivability / the EVADE side (both forces).** Effectors carry `arm_sead_vulnerability` and `emcon`: an always-emitting long-range SAM (S-400, HQ-9) is a SEAD/ARM magnet. Red strike-UAS are modelled with a survivability profile (`can_evade`): the Shahed's "kill chain" is economic saturation + low-altitude penetration, not a clever sensor. Defeat it with cheap layers, not $4M rounds. **Extending the effector dataset:** fill `EffectorSystem` completely — tier, effect, full `envelope` (incl. no-escape zone and altitude band), Pk, magazine (∞ + note for DE), cost-per-shot, `defeats`/`cannot_defeat`, mobility, ARM vulnerability, EMCON. Link `cueing_radar_ids` to the radar layer so the kill chain connects. Mark any dynamic-sim field (`intercept_speed_mach`, `reload_min`) — the static layer ignores time, but the data is there for the future sim.