Radar and ARPA

Prerequisites

This lesson takes radar from the yacht level to the standard of an Officer of the Watch (Unlimited) under STCW Regulation II/1. Before starting you should:

• Know the steering and sailing rules, especially Rules 5, 6, 7, 8 and 19, and the lights and signals lesson.
• Understand true and magnetic bearings, the difference between heading and course over ground, and the effect of a tidal stream on a vessel's track.
• Have used a radar on a yacht or in a simulator: gain, range, EBL and VRM.
• Be working towards, or hold, the STCW Navigation, Radar and ARPA Simulator (NARAS) Operational course required for the certificate (see MSN 1856), which follows IMO Model Course 1.07.

The MCA oral will expect you to plot by hand, explain every number on an ARPA display, and say exactly when not to trust it.

Learning Objectives

By the end of this lesson you will be able to:

• Explain how a marine radar measures range and bearing, and compare X band and S band.
• Describe the operator controls and how to set them for best detection.
• Distinguish relative motion and true motion, head-up, north-up and course-up, and sea and ground stabilisation, and choose the right one for the job.
• Construct a relative plot by hand and from it find CPA, TCPA, and the target's true course, speed and aspect.
• Predict the effect of an alteration of course or speed by own ship.
• Describe ARPA and ATA functions, including acquisition, tracking, vectors, past positions, alarms and trial manoeuvre.
• Explain the main errors and limitations of radar and ARPA.
• Use parallel indexing to monitor the ship's position against a planned track.

How Radar Works

A radar transmits short pulses of microwave energy from a rotating antenna and listens for echoes. Because the pulses travel at the speed of light (about 300 m per microsecond), the time for the echo to return gives the range:

Range = (speed of light × time) ÷ 2

An echo returning after 12.35 microseconds is 1 nm away. The bearing is the direction the antenna was pointing when the echo came back.

Pulse length and PRF

• Pulse length is the duration of each pulse. A short pulse (about 0.05 microseconds) gives good range discrimination, the ability to separate two targets on the same bearing at slightly different ranges. Range discrimination is roughly half the pulse length expressed as distance: 0.1 microsecond gives about 15 m. A long pulse (up to about 1 microsecond) puts more energy on the target and improves detection at long range, at the cost of discrimination.
• Pulse repetition frequency (PRF) is the number of pulses per second. Short pulses go with high PRF on short ranges; long pulses with low PRF on long ranges, giving each echo time to return before the next pulse.
• Minimum range is limited by pulse length and by the time the receiver needs to recover after transmission.

Beam width and bearing discrimination

The horizontal beam width of a ship's X band scanner is typically about 1° to 2°. Two targets at the same range closer together in bearing than the beam width merge into one echo. A longer antenna gives a narrower beam and better bearing discrimination. The vertical beam is wide (about 20° to 25°) so that the ship's rolling does not lose targets.

Radar horizon

Radar waves bend slightly over the horizon, further than light. With heights in metres, the radar horizon in nautical miles is approximately:

Range (nm) = 2.2 × (√ antenna height + √ target height)

With a scanner 36 m above the sea and a target 9 m high: 2.2 × (6 + 3) = 19.8 nm. Low-lying coasts, small craft and ice will be detected far closer. Abnormal refraction changes this: super-refraction (warm dry air over cold sea) extends detection ranges, sometimes producing echoes of land well beyond the normal horizon, while sub-refraction (cold air over warm sea) shortens them.

X band and S band

SOLAS Chapter V Regulation 19 requires ships of 300 GT and upwards to carry a 9 GHz (X band) radar, and ships of 3,000 GT and upwards a second radar, a 3 GHz (S band) or, where appropriate, a second 9 GHz set, functionally independent of the first. Check SOLAS V/19 for the full carriage table, including the tracking aids required by tonnage.

Sensible practice is to run both: the S band set on a longer range for early warning of ships, the X band set on a shorter range for detail and small craft. In heavy rain or a big swell, the S band will often see a ship the X band has lost in clutter.

Operator Controls

Set up the radar in this order after it has warmed up:

1. Brilliance: adjust so that the display can be seen clearly without glare. In darkness turn it down; too bright a picture hides weak echoes.
2. Range scale: choose a range that matches the situation; in open sea a long range (12 nm on one set) for early warning and a shorter range (6 nm) on the other.
3. Pulse length: usually selected automatically with range, but can be changed. Use long pulse on longer ranges or for poor targets at range.
4. Tune: on older sets, adjust for maximum echo strength from a weak target or for the strongest clutter. Most modern sets tune automatically; check it.
5. Gain: increase until a light speckle of background noise just appears. Too little gain loses weak targets; too much floods the screen.
6. Sea clutter (anti-clutter sea, or STC): reduces the gain at short range to suppress sea returns. Use the least that lets you see targets inside the clutter. Too much will remove small boats completely. Automatic clutter is useful but must be checked against manual settings.
7. Rain clutter (anti-clutter rain, or FTC): differentiates the echo so that only its leading edge shows, breaking up large areas of rain. It also weakens real targets, so use sparingly and turn off after the rain passes.
8. Presentation and stabilisation: see below.
9. Performance monitor: check it daily and record the result. A gradually fading monitor indicates loss of transmitter power or receiver sensitivity.

Measuring tools:

• EBL (electronic bearing line): for bearings of targets. Check the heading marker against the gyro and the dead-ahead line regularly; a heading marker error rotates every bearing.
• VRM (variable range marker): for ranges. More accurate than estimating against fixed range rings.
• Parallel index lines: fixed lines on the display for monitoring passing distances (later in this lesson).
• Off-centring: moves own ship off the centre to look further ahead. Do not leave a large off-centre on and forget the targets now behind or abeam with short warning.

Presentation and Motion

Orientation

• Head-up: own heading at the top. Unstabilised. Easy to relate to the view out of the window, but every echo smears whenever the ship yaws, and bearings are relative.
• North-up: north at the top, stabilised by the gyro. Easy to compare with the chart; echoes stay sharp as the ship yaws.
• Course-up: the intended course at the top, stabilised by the gyro. A compromise for coastal work. Resets when course is changed.

Relative motion and true motion

In relative motion (RM), own ship stays fixed at the centre and every target moves according to its motion relative to you. A stationary buoy moves down the screen at your speed on the reciprocal of your course. The trail of a moving target is its relative track: extend it past the centre and you see directly how close it will come.

In true motion (TM), own ship moves across the screen at its true course and speed, and targets move according to their own true motion. Stationary objects stay still. The trail of a moving target shows its true course, which tells you its aspect at a glance.

The diagram below shows the same moment in both presentations, with own ship steering 000° at 12 knots and another ship steering 270° at 12 knots.

Click to zoom

Neither is "better". RM shows the risk of collision; TM shows what other vessels are actually doing. Modern ARPA lets you show RM display with true vectors or TM display with relative vectors. Many OOWs keep a relative motion display with true trails and switch vectors between true and relative to answer two separate questions: "will it hit me?" (relative) and "what is it doing?" (true).

Sea and ground stabilisation

True motion and true vectors need own ship's motion as an input. There are two choices:

• Sea stabilised: own ship's heading from the gyro and speed through the water from a water-track (single-axis) log. Own ship and targets are shown moving through the water. A target's true vector is its heading and speed through the water, so its aspect is correct. A buoy in a tidal stream appears to move against the stream.
• Ground stabilised: own ship's course and speed over the ground, from GNSS or a bottom-track Doppler log. Vectors are tracks over the ground. Stationary objects show no vector, which is ideal for navigation and pilotage.

Click to zoom

For collision avoidance, sea stabilisation is preferred because the COLREGs depend on aspect: which light the other vessel would see, whether she is overtaking, head-on or crossing. In a strong cross stream, ground stabilisation can give a wrong aspect. Picture a vessel steering 000° through the water in a 3 knot easterly set. Ground stabilised, her vector points to about 015°. If you are approaching her from the south-east, the ground vector may suggest you are crossing when in fact you are almost overtaking her.

For navigation, ground stabilisation is better: a ground-stabilised true motion display shows the coast still and your own ship's real track towards it.

Know which stabilisation your ARPA is using. It is shown on the display. If the speed input is wrong, every true vector is wrong.

Relative Plotting by Hand

Every OOW must be able to plot without ARPA. The method uses three points on a reflection plotter, a plotting sheet, or a sheet of paper.

• O: the target's first plotted position.
• A: the target's position after the plotting interval, usually 6 or 12 minutes.
• W: from O, lay off own ship's course and the distance run in the interval, backwards. W is where the target would have been if it were stationary.

Then:

• OA is the target's relative motion. Extend it past the centre to find the CPA.
• WO is own ship's true motion.
• WA is the target's true course and speed.

The triangle shows the vector equation: target's true motion = own ship's motion + relative motion.

Worked plot

You are steering 000° at 12 knots in restricted visibility. You observe a target:

The bearing is almost steady and the range is closing fast: risk of collision exists (Rule 7).

Click to zoom

Step 1: relative motion. O (0000) to A (0012) is 4.0 nm in 12 minutes in direction 224°. Relative speed = 4.0 × 5 = 20 knots.

Step 2: CPA. Extend OA past the centre. The perpendicular from own ship to the extended line is the CPA: 0.2 nm. This is a collision situation.

Step 3: TCPA. From A to the CPA point along the relative track is about 6.0 nm. At 20 knots that is 18 minutes, so TCPA is 0030.

Step 4: true course and speed. Lay off WO: 12 minutes at 12 knots is 2.4 nm on 000°. Put W 2.4 nm due south of O. WA measures 2.8 nm in direction 260°. In 12 minutes that is 14 knots. Target's true course 260°, speed 14 knots.

Step 5: aspect. The bearing of own ship from the target is 045.9 + 180 = 225.9°. Her heading is 260°. You are 34° on her port bow: aspect red 34°. If visibility allowed, you would see her red sidelight and she would see your green.

Step 6: what the rules say. You are in restricted visibility, so Rule 19 applies, not Rules 11 to 18. The target is forward of your beam and not being overtaken, so Rule 19(d)(i) says avoid an alteration of course to port. A substantial alteration to starboard, a reduction of speed, or both, are the options.

Effect of own ship's alteration

To find the effect of a new course or speed, keep the target's true vector WA fixed and draw a new own-ship vector from W. In practice, using the 12-minute triangle:

1. From W, lay off the new own-ship vector for 12 minutes (new course, new speed): call its end O'.
2. O'A is the new relative motion direction. Draw a line through the target's current position (A) parallel to O'A.
3. The perpendicular from the centre to that line is the new CPA.

Results for this target, if the action is taken at 0012 and assuming instant effect:

What this shows:

• Small alterations do little. A 30° alteration gives only 1.2 nm in a situation with a 20-knot closing speed. Rule 8 requires action to be large enough to be readily apparent to another vessel observing by radar. On a radar-only contact, 60° or more is usually needed.
• Speed reduction on a big ship is slow. The figures assume instant effect. A loaded VLCC takes many minutes to lose half her speed. If you use speed, start early and combine it with an alteration.
• Check other targets before you turn. A turn to starboard can create a new close-quarters situation with something on your starboard beam.
• Plan the return. Do not resume course until the target is past and clear (Rule 8(d)). Watch the bearing open and the range start to increase.

ARPA and ATA

An automatic radar plotting aid (ARPA) acquires and tracks targets and calculates the same quantities as a manual plot. IMO performance standards (MSC.192(79), for radars installed from 2008) merged the earlier ARPA and ATA standards: a modern radar's automatic tracking aid (ATA) is what was traditionally called ARPA. The number of targets that must be tracked depends on ship size; check MSC.192(79) for the current table.

Functions

• Acquisition: manual (the operator selects targets) or automatic (within acquisition zones). Automatic acquisition in a busy area or in clutter can acquire false targets and fill the screen; manual acquisition of significant targets is often more reliable.
• Tracking: once acquired, the system updates the target every scan. It shows a trend within about one minute and a reliable prediction within about three minutes of steady tracking. Never base a decision on the first minute of a track.
• Vectors: true or relative, with adjustable length in minutes. A 6-minute vector shows where the target will be in 6 minutes.
• Past positions: dots at regular intervals showing where the target has been. Equally spaced past positions in a straight line show a steady target; curved or uneven spacing shows a manoeuvre.
• Data: for a selected target, bearing, range, course, speed, CPA, TCPA, BCR and BCT (bow crossing range and time).
• Alarms: CPA/TCPA limits, new target in a guard zone, lost target, and system failures. Set CPA/TCPA limits to match the master's standing orders, for example 1 nm and 12 minutes in open sea.
• Trial manoeuvre: simulates a proposed alteration of course and/or speed, with a delay time to allow for the time it takes to start the manoeuvre, and shows the resulting CPAs for all tracked targets. It is the ARPA version of the plotting exercise above. Always set a realistic delay; always check all targets, not only the one you are worried about.
• AIS target integration: AIS targets can be shown on the radar and associated with radar targets. The display must indicate which source is being used. Treat AIS data as supplementary; the radar track is your measurement.

Reading the display carefully

An ARPA tells you what it calculated from the inputs it had. Before you act on a CPA:

1. Is the target being tracked steadily (stable past positions, no "trial" or "acquiring" status)?
2. Which stabilisation is in use, and are the gyro and log inputs correct?
3. Do the vector and past positions agree with the bearing trend you have taken visually or with the EBL?
4. Has the target or own ship just altered? If so, the data lags.

Errors and Limitations

Radar picture limitations

• Shadow and blind sectors: masts, funnels, cranes and containers block the beam. A blind sector astern or on the bow can hide a small vessel completely. Know your ship's blind sectors (they should be posted on the bridge) and alter course slightly from time to time to look into them.
• Side lobe echoes: from a strong target at short range, false echoes appear on an arc either side of the true echo, at the same range.
• Multiple echoes: when a large target is close abeam, the pulse bounces back and forth, giving a series of echoes at multiples of the true range on the same bearing.
• Indirect (false) echoes: reflection off the ship's own structure (a funnel or mast) sends the beam in another direction; the echo appears on the bearing of the obstruction at the correct range of the real object.
• Second-trace echoes: in super-refraction, an echo from beyond the maximum range for the PRF returns after the next pulse has gone and appears at a much shorter range. Changing the range scale (and so the PRF) makes it jump or disappear.
• Interference: from other radars on the same band, seen as curved spokes. The interference rejection control removes it but can also remove weak real echoes.
• Small targets: a GRP yacht, a wooden fishing boat, a RIB or a growler may return a weak echo that is lost in sea clutter or rain. MGN 369 is clear that radar cannot be relied on to detect all small craft.

ARPA errors

• Gyro error rotates every true vector and, on a north-up display, every bearing. A gyro reading 2° low makes every true course 2° wrong.
• Log error makes every true vector wrong in a sea-stabilised display. A log over-reading by 2 knots gives targets a false component of 2 knots in the direction of your own course.
• Target swap: when two targets pass close together, the tracker may jump from one to the other and give nonsense data for both.
• Lost targets: in clutter or in a blind sector, tracking stops. The lost target alarm is easy to ignore in a busy strait; do not.
• Lag after a manoeuvre: when either vessel alters, ARPA needs one to three minutes to settle. During that time the vectors still point the old way. If you see a change of aspect visually before ARPA shows it, believe your eyes.
• Assumed constant motion: every prediction assumes the target holds course and speed. It cannot predict a fishing vessel's next turn or a ship following a bend in a traffic scheme.

Scanty radar information

Rule 7(c) forbids assumptions on the basis of scanty information, especially scanty radar information. A single observation, a CPA calculated in the first minute of tracking, or AIS data alone is scanty. Rule 7(b) requires proper use of radar, including long-range scanning to obtain early warning of risk of collision and radar plotting or equivalent systematic observation of detected objects.

Parallel Indexing

Parallel indexing (PI) is a radar technique for monitoring the ship's position relative to the planned track continuously, without plotting fixes. It works on any radar, in any visibility, and is independent of GNSS.

The principle

When you steer a steady course, a fixed object moves past you on a relative track that is parallel to your course and opposite in direction. If you draw a line on the radar parallel to your course, at the planned passing distance from a radar-conspicuous object, the echo of that object should slide along the line as you pass it.

Click to zoom

Setting up a PI line

1. In the passage plan, choose a radar-conspicuous object near the leg: a steep headland, a small island, a lighthouse structure or a racon. Avoid low shelving coasts and buoys that can drag.
2. Measure on the chart the perpendicular distance from the planned track to the object: the cross index range (here 1.5 nm to starboard).
3. Draw the PI line on the radar parallel to the course (000°), 1.5 nm to starboard. On a north-up or course-up stabilised display, it stays correct as the ship yaws. On head-up it does not; use a stabilised display.
4. Mark any wheel-over points or limits along the line. A cross mark on the PI line at the range where you must turn gives a wheel-over position independent of GNSS.

Reading it

• Echo slides down the line: you are on track.
• Echo inside the line (closer to own ship): you are closer to the object than planned, being set towards it, or steering wrong. Correct now.
• Echo outside the line: you are further off than planned.

Because you are watching continuously, you see a set the moment it starts, long before a three-minute fix interval would show it. In pilotage waters, PI lines are the standard primary method of monitoring on many ships, with ECDIS and GNSS as the check.

Precautions

• Draw the PI line parallel to the planned ground track, not to the course steered. When you steer into a cross stream, a fixed object still moves past you parallel to your track over the ground, which is exactly what you want to monitor.
• Check the radar's range accuracy and heading marker alignment before relying on fine PI margins.
• Make sure the object is identified correctly. A mistaken headland gives perfect-looking but wrong guidance.

Worked Example: Night Watch in Rain

Situation: You are OOW on a 230 m container ship, course 210°, 18 knots, in the English Channel at 0300. Heavy rain showers are moving through. The X band radar on 6 nm range shows a large rain cell ahead; the S band on 12 nm shows two targets.

Step 1: set up. You reduce rain clutter on the X band until the cell breaks up, accept that it is still poor inside, and rely on the S band for the area of rain. You confirm both sets are north-up, sea stabilised, with the log reading 18.2 knots through the water against a GNSS SOG of 17.1 knots: a 1 knot foul stream, as expected.

Step 2: acquire. You manually acquire both S band targets. After three minutes of steady tracking, Target 1, bearing 200° at 9 nm, shows true course 030°, speed 12 knots, CPA 0.3 nm, TCPA 18 minutes. Target 2, bearing 235° at 7 nm, shows true course 120°, speed 10 knots, CPA 2.5 nm passing ahead.

Step 3: assess. Target 1 is nearly a reciprocal course: a head-on situation if in sight, and in restricted visibility (rain) a close-quarters risk under Rule 19. Visibility in the showers is under 1 nm.

Step 4: trial manoeuvre. You set a trial alteration to 260° with a 1-minute delay. ARPA predicts CPA 2.6 nm for Target 1, but Target 2 now has a CPA of 0.6 nm because you are turning towards her. You try 240°: Target 1 CPA 1.8 nm, Target 2 CPA 1.4 nm. You also consider reducing to 12 knots with an alteration to 240°: Target 1 2.0 nm, Target 2 1.6 nm.

Step 5: act. You call the master as required by standing orders (restricted visibility, CPA under 1 nm), sound one prolonged blast at intervals of not more than 2 minutes, put the engine on standby, and alter to 240° early and decisively. You watch the past positions of both targets and the bearing trend on the EBL.

Step 6: monitor and resume. When Target 1 is past and clear on your port quarter and opening, and Target 2 is clear ahead, you return to 210° gradually, checking that no new target has appeared in the rain.

Common Mistakes

• Trusting the first minute of ARPA data. Wait for a stable track, and cross-check with compass bearings.
• Using ground-stabilised vectors to judge aspect in a strong stream. Switch to sea stabilisation for collision avoidance and check the log input.
• Too much sea clutter. It removes small craft. Use the least that works, and vary it to look into the clutter.
• Running one radar only. Use both, on different ranges, especially in rain.
• Small alterations. A 10° or 20° course change on a radar-only contact is not readily apparent. Make it substantial.
• Turning to port for a target forward of the beam in restricted visibility. Rule 19(d)(i) says avoid it, except for a vessel being overtaken.
• Forgetting blind sectors. A yacht can sit in the blind sector of the foremast for many minutes.
• Ignoring trial manoeuvre on other targets. An alteration that solves one problem often creates another.
• Leaving off-centre or a long range on for too long. Targets close abeam and astern can be missed.
• Treating AIS as radar. AIS shows only vessels that transmit, with data that may be wrong. Radar sees what is actually there, within its limits.

Summary

• Range comes from echo time; bearing from antenna direction. Short pulse gives good range discrimination; narrow beam gives good bearing discrimination.
• X band: sharper and better on small targets, but more affected by rain and sea clutter; triggers SARTs. S band: better in rain and heavy weather.
• Set brilliance, range, gain, sea and rain clutter carefully; check the performance monitor and heading marker.
• Relative motion answers "will it hit me?"; true motion answers "what is it doing?". Sea stabilisation for collision avoidance; ground stabilisation for navigation.
• In the O-A-W plot, OA is relative motion, WO own ship, WA the target's true course and speed. Extend OA for the CPA.
• Alterations must be large; check every target before turning; speed reduction on a big ship is slow.
• ARPA needs about three minutes of steady tracking and is only as good as its gyro and log inputs.
• Parallel indexing monitors the track continuously against a radar-conspicuous object, independent of GNSS.

Check Your Understanding

1. Why would you choose the S band radar to look for a ship in heavy rain, and the X band radar to look for a SART?

Answer: S band (about 10 cm wavelength) is much less attenuated and cluttered by rain than X band, so it detects ships through rain cells better. A radar SART responds only to 9 GHz (X band) radar pulses, so only the X band set will show its line of 12 dots.

1. What is the radar horizon for a scanner 25 m above the sea detecting a target 16 m high?

Answer: 2.2 × (√25 + √16) = 2.2 × (5 + 4) = 19.8 nm, in normal propagation conditions.

1. In a relative plot, what do the lines OA, WO and WA represent?

Answer: OA is the target's relative motion (direction and rate relative to own ship); WO is own ship's true course and distance run in the plotting interval; WA is the target's true course and distance run, from which its true speed is found.

1. A target's past positions are equally spaced and in a straight line, but its ARPA vector suddenly swings. What may have happened?

Answer: Either the target has started to manoeuvre (watch for uneven spacing of new past positions), own ship has altered course or speed, or there may be a target swap or input error such as a gyro or log fault. Cross-check with compass bearings and the other radar before acting.

1. Why is sea stabilisation preferred for collision avoidance?

Answer: Sea-stabilised true vectors show the target's heading and speed through the water, which gives its true aspect and lets you apply the COLREGs (head-on, crossing, overtaking) correctly. Ground-stabilised vectors show tracks over the ground, which in a cross stream can differ significantly from heading and mislead you about aspect.

1. Own ship 000° at 12 knots. A target's true course is 260° at 14 knots and it is closing with a CPA of 0.2 nm. Why might a 30° alteration to starboard be inadequate?

Answer: Because the closing speed is about 20 knots, a 30° alteration only opens the CPA to about 1.2 nm. It may also not be readily apparent to the other vessel observing by radar, as Rule 8(b) requires. An alteration of 60° or more (giving about 2.5 nm) is more appropriate.

1. List four limitations of ARPA that an OOW must remember.

Answer: Any four of: data takes about one to three minutes to stabilise after acquisition or after either vessel manoeuvres; errors in gyro or log input make true vectors wrong; target swap when targets pass close together; loss of tracking in clutter or blind sectors; predictions assume the target holds its course and speed; automatic acquisition may miss small targets or acquire false ones.

1. You are passing a steep island which should be 2 nm to port on a parallel index line. The echo is moving down the screen 1.6 nm from the centre. What does this mean and what do you do?

Answer: The echo is inside the PI line, so the ship is 0.4 nm closer to the island than planned, probably being set towards it. Alter course to starboard to regain the track, check your position by another method (GNSS, ECDIS, a radar fix), and allow for the set on the next leg.

1. Name three types of false echo and how to recognise one of them.

Answer: Side lobe echoes, multiple echoes, indirect (false) echoes and second-trace echoes. For example, multiple echoes appear as a line of equally spaced echoes on the same bearing at multiples of the true range when a large ship is close abeam; second-trace echoes jump or vanish when the range scale (and so the PRF) is changed.

Further Reading

• A.G. Bole, A. Wall and A. Norris, Radar and ARPA Manual, 3rd ed.
• IMO Resolution MSC.192(79), radar (and ARPA/ATA) performance standards
• IMO Model Course 1.07, Radar Navigation at Operational Level
• MGN 369 (M+F) Amendment 1, navigation practices in restricted visibility
• MGN 315 (M), Keeping a safe navigational watch on merchant vessels
• A.N. Cockcroft and J.N.F. Lameijer, A Guide to the Collision Avoidance Rules, 7th ed.
• SOLAS consolidated edition (IMO), Chapter V Safety of Navigation
• ICS Bridge Procedures Guide, 6th edition