Resolution is whether a radar sees two aircraft flying close together as two. A wide beam or a long pulse makes them merge into a single blip.
In practiceA formation of four aircraft can appear on the display as a single target at long range. The air defence has to estimate the number of aircraft in an attack (the raid count), and attackers deliberately fly close together to disguise it.
Resolution is a radar's ability to see two targets that are close together as two targets rather than one. Resolution is not the same as accuracy: accuracy says how well you measure the position of one target, resolution whether you can separate two targets. Radar resolution has three dimensions: angle, range and velocity.
Angular resolution
Angular resolution is determined by the width of the antenna beam: two targets at the same range are only separated if they are further apart in angle than about one beamwidth. The beamwidth follows from the wavelength and the antenna size:
X-band (λ = 3 cm) with a 1 m dish: θ ≈ 70 × 0.03/1 ≈ 2.1°. At 50 km the beam is then ~1.8 km wide — two aircraft flying 1 km apart there merge into one blip. The same resolution at VHF (λ = 2 m) would require an antenna of almost 70 m — that is why low-frequency radars measure so coarsely.
The cross-range resolution (in metres) grows linearly with range: ΔX ≈ R × θ. Angular resolution is therefore mainly the limiting factor at long range. This explains the entire frequency classification of Military applications within the EMS: precise aiming requires high frequencies or large antennas.
Range resolution
Range resolution is the minimum distance in depth at which two targets are still separated. Two targets merge when their echoes overlap in time — and that happens when they are less than half a pulse length apart:
Note that range resolution does not depend on range: a resolution of 150 m is 150 m, at 5 km and at 300 km. That is exactly the opposite of angular resolution, which deteriorates with range. At long range, range is therefore a radar's "sharpest" dimension.
High range resolution is about more than separating aircraft flying in formation: with resolutions of a few metres or less a radar can build a profile of a target (which parts reflect where) and so help with classification and identification — and with distinguishing a real target from chaff or a decoy.
Velocity resolution
Velocity resolution is the ability to separate two targets with (almost) the same position but different velocities — the domain of the Doppler filter bank of a pulse-Doppler radar (Pulse-Doppler radar and Moving Target Indicator). The deciding factor is the observation time (dwell time): the longer the radar looks at the target coherently, the more finely the Doppler frequencies can be pulled apart:
Here physics collides with operations: looking longer per direction means scanning more slowly — a multifunction radar has to divide its time between measuring velocity finely and keeping the air picture fresh.
Velocity resolution has an important EW application: separating a target from its surroundings and from false targets. Chaff clouds slow down almost immediately to wind speed; an aircraft does not. A radar with good velocity resolution sees that difference within seconds and can "see through" the chaff. The same applies to recognising slow-moving decoys among real targets.
The resolution cell
The three resolutions above (angle, range, velocity) are not separate numbers — together they span the resolution cell: the smallest "block" of space (and velocity) that the radar cannot subdivide further. Two targets that fall within one resolution cell merge into one blip. Only if they are far enough apart in at least one dimension (range, azimuth, elevation or Doppler) does the radar see two.
The dimensions of the cell:
- Depth (range): ΔR = c/(2B) — constant, independent of range (Range resolution).
- Width (cross-range/azimuth): R × θaz — grows with range (Angular resolution).
- Height (elevation): R × θel — also grows with range.
Why this matters:
- Target separation: a small cell separates aircraft in formation, ships in convoy and a target with a decoy (Decoys); a large cell lumps them together.
- Clutter: the larger the cell, the more ground clutter (Ground clutter) fits into it and the lower the signal-to-clutter ratio.
- Classification: very small cells (high resolution in all dimensions) make a "profile" of a target possible — the beginning of target recognition (NCTR).
Pulse-Doppler radars add a fourth dimension: the Doppler bin (Velocity resolution). Two targets in the same range-azimuth-elevation cell can still be separated if their velocities differ enough — which is exactly why the notch manoeuvre (Pulse-Doppler radar and Moving Target Indicator) works: equal to the clutter in position and in Doppler, and therefore invisible.
Resolution is not the same as accuracy
Resolution and accuracy are often confused, but they answer different questions:
- Resolution: can the radar separate two targets? (Angular resolution to The resolution cell)
- Accuracy: how close is the measured position of one target to its true position?
Accuracy can be much better than resolution. A monopulse radar (Cassegrain, monopulse feed and duplexer) with a 2° beam often measures the bearing of a single target to within a tenth of a degree, because it does not measure the beamwidth but the centre of the echo. The higher the SNR, the more accurate: the measurement error decreases roughly with the square root of the SNR.
The remaining errors fall into three groups: random errors (noise, the fluctuating echo), systematic errors (an antenna not exactly aligned to north, a delay in the cables) and propagation errors (refraction, multipath over the sea). Systematic errors are treacherous: they do not disappear by measuring more often, only by calibrating.
A tracking radar that follows a target accurately is not necessarily able to separate two aircraft flying close together. Formation flying exploits this, and so does glint: the shifting reflection points of a large target, or of two targets in one resolution cell, make the measured centre jump around — something a cross-eye jammer (Deceptive jamming) deliberately imitates.