Home Harpoon II & Harpoon 3 The Harpoon 3 Sonar Model
 -

# The Harpoon 3 Sonar Model

 -
 Theory

This chapter explains the theory of sonar only as far as it is necessary to understand the physical background of the formulas being specified below. It is far from being an exhaustive treatise on marine acoustics; readers who are interested in further information on this subject have to look for other, more specific sources. Basic knowledge in math and physics is recommended.

Sound Measurement

The Harpoon 3 game engine uses a logarithmic scale for sound intensities, so the author uses in his formulas the measure "dB". However, the Decibel scale normally relates on special reference intensities (which vary between the countries). But in H3 only differences between given noise levels are important, so this information is irrelevant anyway. The only noise levels declared by the game engine are the background noise levels, all others are given in the database.

I  =  p2 / ( c * ro )         [N/(m*s)]

p  :  sound pressure
c  :  speed of sound (in water ~1400m/s)
ro :  specific gravity (water ~1000kg/m^3)

As the sound intensity can extremely vary (up to 1014), it is converted into the logarithmic noise level L:

L  =  10 * log10( I / I0 )         [dB]

I0 :  reference intensity (normally with p0 = 10-6 Pa at 1m)

This scaling is handy but has to be used with caution. Some examples for the relation between sound intensity and noise level:

 2 * I L + 3.0dB 3 * I L + 4.8dB 5 * I L + 7.0dB 7 * I L + 8.5dB 10 * I L + 10.0dB

Noise Signature by Vessels

Every ship or submarine emits sound energy which is unevenly spread across the whole acoustic spectrum. The frequency allocation of the emitted noise is unique for each vessel. To reduce calculation time Harpoon 3 only works with discrete bands which add up the sound intensities within large frequencies ranges to few total band noise levels. There are noise levels for three frequency bands: L(ow)F, M(edium)F, and H(igh)F. The Harpoon 3 database characterizes units with the average of these three values (Passive Sonar Cross-Section). The total noise levels are not the same for all frequency bands. To reflect this Harpoon 3 calculates the band noise with the average Passive Sonar Cross-Section by adding a band-depending constant:

L(xF)  =  Lpscs  +  Gf         [dB]

Like the steady frequency allocation their proportion to each other should be unique for each ship/sub class, too. Unfortunately, the Harpoon 3 sonar model only knows a standard frequency allocation being used for all units (including weapons). The noise intensity emitted by a vessel strongly depends on its speed. Harpoon 3 works with a simple, linear rule:

L(v)  =  L0  +  mv * v  +  Gc         [dB]

L0 is the LF/MF/HF total band noise level (see above). The noise increase factor mv is the same for all three frequency bands. Cavitation which occurs when the ship/sub runs on full or flank speed is reflected with the constant Gc. This rule means that the Passive Sonar Cross Section in the Harpoon 3 database is something like an "idling noise level". The max range for direct-path detection is 15nm.

Passive Sonar Background Noise

Sonar have to find the sound signature of a target within the background noise. This noise is produced by the sonar itself (reflected at the sensor sensitiveness), the own vessel, and the environment. As it (primarily) depends on the noise level ratio of target signature to interference whether a sub/ship is detected, this phenomenon must be reflected. Sound emitted by a ship/submarine is not only heard by other units; acoustic devices on the emitter itself receive it, too. This noise is calculated by Harpoon 3 roughly as the target noise level. Oceans are not very silent even without any shipping in. Harpoon 3 works with an ambient noise level calculated like this:

La  =  La,0  +  mssl * (Sea State Level)         [dB]

La,0 represents the noise produced by geological phenomena, animals etc. There is a certain La,0 for each Frequency band. Waves increase the noise level by mssl for each Sea State Level. Ambient noise and noise by the own ship/submarine effect together on acoustic devices, so their noise levels have to be added. Because additions are impossible in the logarithmic scale, the total sensor background noise has to be calculated with the sound intensities instead of the noise levels:

Lb  =  10 * log10( Ia  +  Ir )         with     Ia/r  =  10^( La/r / 10 )

Sound Propagation Characteristics

On its way from the emitter to the receiver, sound is reduced by two phenomena: propagation and absorption. Propagating sound forms a sphere with the emitter at the center of it. As the area of this sphere grows by the squared radius, the sound intensity decreases in the same way:

Iprop  =  I / R2
Lprop  =  10 * log10{ I / ( I0 * R2 ) }  =  L  -  20 * log10( R )         [dB]

The radius R is given in nautical miles, so the reference distance where the sound intensities are measured would actually be 1nm. But the reference distance for all entries in the database is 1m. Therefore the noise level first has to be re-scaled:

Lnm  =  Lm  -  20 * log10( 1852 )  =  Lm  -  65.35 lsi         [dB]

Propagation is identical for all frequencies. Additionally, the medium in which the sound waves run (here water) absorb sound energy. This effect strongly varies with the frequency: High frequencies are absorbed really fast, low frequencies can run hundreds of miles. For all frequencies the rule is that the noise level decreases by a constant number of Decibels for every mile:

Labsorb  =  L  -  md * R         [dB]

The value md determines how fast the sound energy is absorbed by the water. Harpoon 3 knows three different reduction factors, one for each frequency band.
Both effects are together reflected in Harpoon 3. As a simple rule, low-frequent noise mostly decreases because of propagation, while high-frequent sound energy is primarily absorbed by water.

Active Sonar

From the physical aspect active sonar in general works like passive sonar. The only difference is that the searching unit first has to produce the sound it wants to receive from the target. So the sound suffers the same losses by propagation and by absorption like at passive sonar, but with the double range - from the sonar to the target and back to the searching vessel:

Iprop  =  ( I / R2 ) / R2  =  I / R4
Lprop  =  10 * log10{ I / ( I0 * R4 ) }  =  L  -  40 * log10( R )         [dB]

Labsorb  =  L  -  md * 2 * R         [dB]

To compensate for this fact active sonar emit sound at much higher energy than any existing vessel could do. Active and passive sonar differ in terms of background noise. As all active sonar being in service are still narrow-band systems, the background noise produced by both the environment and the own vessel is less than at passive systems with their wide processed spectrum. For active sonar Harpoon 3 does not reflect the self-emitted noise like like it does at passive sonar; this source of interference is ignored. But higher speed of the searching unit still reduces the sensor sensitivities because turbulence on the surface of the sonar dome, which grow with speed, distort the emitted and received signals.

The Layer

Harpoon3 has an acoustic layer at 40-50m. Submarines at Intermediate or greater depth are below the layer. Submarines at shallow, periscope or surface depth are above the layer. The layer varies in strength depending on the sea state and is calculated the following way:

layerStrength = min ( 10 - Sea State )

Ships and submarines with Towed Array and VDS systems are not affected by this rule and can make long-range detection below the layer.

Convergence Zones

Convergence zones are simulated by adding 15dB to the target noise level. The size of the CZ is calculated differently depending on where you are in the world. Normally, CZ's are calculated the following way:

CZ radius = 50,000 yards + a random number between 0 and 21,000 yards.

CZ width = 3,000 yards + a random number between 0 and 8,000 yards.

CZ minimum sea depth = 250m

In for example the Mediterranean the CZ's are calculated differently:

CZ radius = 30,000 yards + a random number between 0 and 11,000 yards.

CZ width = 2,000 yards + a random number between 0 and 6,000 yards.

CZ minimum sea depth = 125m

Wave Noise Interference

In shallow waters (depths of 250m or less, 125m in the Med) , the performance of LF sonar is greatly reduced by wave noise interference (target signature is reduced by 8dB). MF sonar is less affected (5dB), while HF sonar hardly affected at all (2dB). A target located in a CZ produces an additional 15dB.

Towed Array Sonar

Towed Array and VDS are most effective when used at speeds from 5 to 13 knots. The systems can not be used at speeds greater than 23 knots. These systems can search below the layer without suffering any penalties.

Dipping Sonar

Dipping sonar on helicopters is automatically deployed when the helicopter is hovering at altitudes of 50 meters or lower. Dipping sonar is not affected by self noise.

Submerged submarine with mast

A submerged submarine with mast up can be detected at up to 5nm. The chance for a detection is calculated this way:

percentageDetect = 30 - ( range * range ) - ( Sea State * 5 )

The Alerted Operator modifier is used for close targets and adds 2dB to the target noise level.

 -
 Formulas

----- Passive Sonar -----

Target Noise
Every target ship/sub in Harpoon 3 emits sound within each frequency band at the noise level Lt:

Lt  =  ( Lt,0  +  Gf)  +  mv,t * vt  +  Gc         [dB]

Lt,0  :    target Passive Sonar Cross-Section from the database
Gf    :   divergence of band noise level from average noise level
LF:   Gf  =  -6 dB         MF:   Gf  =  +5 dB         HF:  Gf  =  -19 dB
mv,t  :   target noise increase by speed
mv,t  =  1 dB/kts
vt     :   target speed
Gc    :   noise increase by cavitation (only at full/flank speed)
Gc  =  2 dB

Ambient Noise
The ambient noise level La is:

La  =  La,0  +  mw,p * ssl         [dB]

La,0  :   ambient noise basis level
LF:   La,0  =  87 dB         MF:  La,0  =  90 dB         HF:   La,0  =  73 dB
mw,p :   noise increase by waves
mw,p  =  5 dB/(sea state level)
ssl     :   sea state level

In Harpoon 3, the ambient noise level does not vary with depth.

Sensor Background Noise
The passive sonar also receives noise from its own carrier with the intensity Lr:

Lr  =  Lr,0  +  mv,r * vr         [dB]

Lr,0  :   receiver Passive Sonar Cross-Section from the database. For Towed Array sonar and VDS this value is one-third the rear Passive Sonar Cross-Section from the database.
mv,r  :   receiver noise increase by speed
LF:   mv,r  =  3.6 dB/kts         MF:   mv,r  =  4.2 dB/kts         HF:   mv,r  =  5.5 dB/kts

Cavitation at full/flank speed has no effect on Lr. The total intensity Lb,p of the sensor background noise then is:

Lb,p  =  ( 10 / mb ) * log10{ 10^( Lr / 10 )  +  10^( La / 10 ) }         [dB]

mb   :   background noise reduction factor
LF:   mb  =  3.4         MF:   mb  =  2.9         HF:   mb  =  3.7

The logarithmic addition of Lr and La can easily be done with the table below; for L1 take the larger one of Lr and La, for L2 take the smaller one of both.

 L1  -  L2 SUM( L1, L2 ) 9 L1 9 L1  +  1 8 L1  +  1 7 L1  +  1 6 L1  +  1 5 L1  +  1 4 L1  +  1 3 L1  +  2 2 L1  +  2 1 L1  +  3 0 L1  +  3

To get Lb,p, the result from the table then only has to be multiplied with 1/mb.

Condition for Passive Detection
A target is detected by passive sonar at the range R with:

Lt  -  { 20 * log10( R )  +  65 dB  +  md * R }  -  Lb,p  =  Gs,p

md   :   dispersion factor
LF:   md  =  1/6 dB/nm         MF:   md  =   1 dB/nm         HF:   md  =  3 dB/nm
Gs,p :   sensor Passive Sensitivity from the database

The dispersion factor (aka Attenuation Coefficient) varies by ocean, season and frequency. Only the Atlantic norms are used in Harpoon 3, and the values are not set by location and time of year. In shallow waters (depths of 250m or less, 125m in the Med), the performance of LF sonar is greatly reduced by wave noise interference (target signature is reduced by 8dB). MF sonar is less affected (5dB), while HF sonar hardly affected at all (2dB). A target located in a CZ produces an additional 15dB of sound.

With LF at all ranges, MF and HF at < 5000 yards:

20 * log10( R )  +  md * R  =  Gd( R )

MF and HF at 5000+ yards:

37 + 10 * log10( R )  +  md * R  =  Gd( R )

and the table below the detection range can be quickly estimated. Values in the table are all rounded down.

 R [nm] LF:   Gd( R ) MF:   Gd( R ) HF:   Gd( R ) 2 6 8 12 3 10 12 18 4 12 16 24 5 14 18 28 6 16 21 33 7 18 23 37 8 19 26 42 9 20 28 46 10 21 30 50 11 22 31 53 12 23 33 57 13 24 35 61 14 25 36 64 15 26 38 68 16 26 40 72 17 27 41 75 18 28 43 79 19 28 44 82 20 29 46 86 21 29 47 89 22 30 48 92 23 31 50 96 24 31 51 99 25 32 52 102 26 32 54 106 27 33 55 109 28 33 56 113 29 34 58 116 30 34 59 119 31 34 60 122 32 35 62 126 33 35 63 129 34 36 64 132 35 36 65 135

Lb,p  =  ( 10 / mb ) * log10{ 10^( Lr / 10 )  +  10^( La / 10 ) }         [dB]

mb   :   background noise reduction factor
LF:   mb  =  3.4         MF:   mb  =  2.9         HF:   mb  =  3.7

----- Active Sonar -----

Sensor Background Noise
The active sonar background noise level Lb,a depends on the receiver speed and the sea state:

Lb,a  =  Lb,a,0  +  ma * vr  +  mw,a * ssl         [dB]

Lb,a,0 :   active sonar basis background noise level
LF:   Lb,a,0  =  34 dB         MF:   Lb,a,0  =  21 dB         HF:   Lb,a,0  =  36 dB
ma     :   sensor background noise increase by speed
ma  =   0.5 dB/kts
mw,a  :   sensor background noise increase by waves
mw,a  =  4 dB/(Sea State Level)
ssl      :   Sea State Level

Cavitation by the receiver does not increase the active sonar background noise.

Condition for Active Detection
A target is detected by active sonar at the range R with:

( Ls,o  +  GA  -  { 40 * log10( R )  +  65 dB  +  md * 2 * R } ) / 2  -  Lb,a  =  Gs,i

Ls,o   :   Sensor Output value
GA     :   Target Active Sonar Cross-Section from the database
GA  =  10 * log10( cross-section area in m^2 )
md     :   dispersion factor
LF:   md   =  1/6 dB/nm         MF:   md  =  1 dB/nm         HF:   md  =  3 dB/nm
Gs,i    :   Sensor Input value

The signal value is halved compared to the passive formula because the signal is traveling twice as far (once out to the target, then the reflected signal comes back), and thus is subject to propagation loss twice. Over the ranges active sonar is effective, this gives very nearly the same results as doing the real calculation and is much faster (remember, the target machine was a 486/100MHz). On reflection, the interference level might not really be halved (since it would get applied twice), but this is how the formula the Harpoon II programmers were given worked out.

 -
 Examples
 -
 Sonar in Real Life

This is a compilation of the tons and tons of information that was exchanged around the time when we designed the sonar model for the DB2000 Harpoon 3 database.

Passive Sonar

Sea state and Target Noise always causes the biggest changes to sonar predictions. Target Noise and Target Speed were always calculated as one entry = Target Self Noise. Likewise, Receiver Self Noise, Receiver Speed, and Cavitation were grouped together as = Own Ship Noise. They were combined because the separate items are directly proportional to each other.

With newer nuclear submarines at speeds below 10-12 knots you're talking about detection ranges of less than a mile. Diesel submarines can become undetectable passively because they can shut down everything that makes noise, regardless of how modern they are. A modern diesel at a 2-3 knot patrol speed is probably not detectable beyond 1000 yards passively, less in high ambient noise environments. Obviously, a lot depends on ambient noise, propagation paths, layer depth, the sensitivity and location of the passive sonar receiver, proficiency of the submarine crew and operating mode, etc. In fact, in an inshore environment (shallow water, high ambient noise, high shipping density, high wreck density), attempting to track a diesel submarine passively is virtually impossible, and extremely difficult actively, and the US Navy relies primarily on non-acoustic methods for initial detection, i.e. a periscope search using ISAR radar being the most effective. MAD in a shallow water environment is handicapped also... wrecks, bottom topography, geologic features, etc., all contribute to false MAD contacts and high magnetic noise, reducing the detection range. For that reason, passive detection range for a diesel submarine in shallow water should be Zero.

Factors limiting active sonar performance in shallow water (the littoral environment) also play a major role... active sonar frequency and power affect bottom reverberation and absorption. Bottom compositions are rated on their ability to absorb and reflect sound energy. A muddy bottom will absorb a lot of energy, whereas a rocky, gravel bottom will reflect and scatter a lot of energy. Again, wrecks will give false contacts. A good diesel sub CO can avoid active detection by going dead in the water and pointing the bow or stern towards the sonar, reducing the target strength by as much as 80 per cent and not providing any Doppler return to the sonar. Or he can bottom, in which case his target echo is masked by the bottom reverb, and if he bottoms near a wreck you've got more problems.

The big point is that the environment pays a major role in the ranges observed. A Victor III in the Norwegian Sea (relatively quiet sea and deep) at 12 knots may be detected at several miles. The same submarine in the Med (relatively shallow and very noisy) may be detected at a 1000 yards. At flank speed (27 knots), the Victor III may be detected at 20 miles direct path, 25-40 miles bottom bounce, and possibly to 3 or more CZ's (convergence zones) at 30-33 miles, 60-66 miles and 90-99 miles in the Norwegian Sea, by ship based sensors and sonobouys, and for literally thousands of miles by SOSUS. What you see here is an overlap of ranges depending on transmission path, and that is entirely normal and expected. SOSUS exploits the deep sound channel, low frequency noise propagated for thousands of miles in a duct created by the effects of pressure and temperature at those depths.

This is just the tip of the iceberg, and my intent was to point out that you can't just assign hard and fast numbers. Under the right conditions a carrier may be detected acoustically well in excess of 140 miles, or may not be identified at all until it's in visual range. Assuming the carrier is detected at 140 miles, can the operator classify it as a carrier? Maybe, maybe not. If he is operating a sophisticated narrowband acoustic processor, possibly, assuming the carrier isn't using acoustic deception. If it is a broadband system (namely an active sonar being used in a passive mode), all he knows is something is making a lot of noise on a given bearing. That, combined with other intelligence may provide another piece of the puzzle, but you can't definitively classify a target with broadband sonar. A carrier launching and recovering aircraft is a different story. The noise of the catapults hitting the water brakes every 30 seconds or so is very distinctive, can be heard for long distances, and any submarine acoustic analyst has probably been trained to recognize that sound. A more comprehensive list of variables:

Source Level (SL) expressed in decibels (dB). Sound pressure level of individual noise sources of the target, i.e. propellors, drive shafts, reduction gears, steam turbines, electrical generators, reactor coolant pumps, diesel engines, main propulsion motors, other pumps and motors, speed related components (hull resonance's occurring at different speeds). ASW tacticians and operators will use the most detectable steady state noise sources for a given target as their primary detection, classification, and tracking frequencies.

Ambient Noise (AN) expressed in decibels at a given frequency(dB)(which includes sea state, rain, biologics, distant shipping noise, underwater geologic disturbances, etc.) i.e., anything not target related.

Recognition Differential (RD) expressed in dB. The sensitivity of the equipment and operator proficiency (i.e. ability to detect and classify a target unalerted). Tends to be a subjective number.

Directivity Index (DI) in dB. The improved sensitivity of directional sonar systems, where the receivers can be focused on a given sector.

Propagation Loss (PL) in dB at a given freq.. Sound energy is attenuated by spreading losses, absorption (sound energy converted to heat energy), reflection, refraction, etc. Prop loss varies directly with frequency.

Self Noise (SN) in dB. Primarily flow noise over the sensor array, but can also include system noise, artifacts (caused by electrical interference within the equipment-- a design limitation, also affects RD).

Target Strength (TS) in dB. The "sonar cross section" of a target. Amount of sound energy reflected from a target.

Signal Excess (SE) in dB. How much signal is left after accounting for all the variables mentioned above.

These variables are what make up the passive and active sonar equations. The passive sonar equation is as follows:

SE = SL - PL - AN - RD + DI

Propagation loss is usually calculated and displayed on a graph, to which we apply a Figure Of merit (FOM), calculated from a version of the passive equation:

FOM = SL - AN - RD + DI

Using this graph we can determine the expected detection range for a given frequency, including the usability of various transmission paths... direct path, bottom bounce, convergence zones.

Active Sonar Equation:

SE = SL - 2PL - AN - RD + TS + DI

2 x Prop loss because sound energy must travel two ways

I guess by now you have learned that sonar performance/predictions in the real world are very unpredictable and therefore almost impossible to duplicate in a game. While new systems improve our ability to process sonar information (it even provided us with some very accurate targeting solutions against enemy surface ships over the horizon) it was still subject to mother natures whims. The following "stuff" is data that you may be able to use to check the accuracy of your sonar model. Don't sweat over it too much because replicating mother natures affect on sonar is difficult.

In simulations of attacks by Class 206/206A subs (Germany) vs. Russian subs (SSK) in the Baltic Sea, first contacts (does not equal a firing solution) were in the range of 3000-5000 yards. Sometimes, first contact was the torpedo launch by the Russian sub. The tracking range for ASW groups and convoys was greater, but we also engaged them at greater distances, sub vs. sub was always a close quarter king of thing. During those simulations (it's actually a real part of a 206 attack center, which gets fed with signals by computers, and an auditorium with a large replay screen for analysis, looks similar to the H2 map display) the German submarines got mostly sunk by two kinds of attacks: one was a torpedo send back down the bearing of our own torpedo, and the second, more effective one, were ASW helos. During real-life exercises with other 206A boats firing solutions were achieved at about 1500 yards. This was in the North Sea, near the Channel entrance.

Towed array sonar is an entirely passive system. Basically, the array is a neutrally buoyant fluid filled tube containing passive hydrophones attached to a mile long tow cable. The depth of the array is controlled by how much tow scope is out, and the speed of the ship. Our currently deployed system is the SQR-19 CATAS, which stands for critical angle towed array system. The earlier SQR-18 array was actually attached to the SQS-35 VDS "fish" or transducer body and the depth of the array was controlled by the depth of the VDS fish. But again, there were some serious maneuvering restrictions. The only maneuvering restrictions for the SQR-19 are that you can't back down, or make more than a 180 degree turn. There are no speed restrictions. The information gained by towed array sonar includes classification information, bearing, and speed of the target... and over time you can get a target fix by maneuvering the array or having more than one sensor in contact to get a cross fix.

A quick blurb on passive sonar: We basically exploit noise from targets two ways, with broadband processing and narrowband processing. Basically, broadband noise is generated by the movement of the hull of a ship or sub through the water. It occurs over a wide frequency range. Hull mounted sonar systems can detect this passive broadband noise and display the data as a noise spike on the bearing to the target... that is all it can tell you. Towed arrays and passive sonobouys use sophisticated spectrum analysis processors to pull out specific frequencies that the target is generating... i.e., from the propellors, drive shafts, reactor pumps, reduction gears, blowers, electrical generators, etc. There are other narrowband sources that a submarine generates that are speed dependent... various hull resonance's at different speeds, for example. This is how we "fingerprint" a specific class of submarine, sometimes down to an individual hull number.

Some of the factors that affect active and passive sonar performance include the temperature profile, the depth, ambient noise, the radiated noise or source level of the target (passive sonar), target strength (similar to radar cross section for an aircraft, for active sonar), own ships noise, propagation loss (sound energy being absorbed, reflected, or refracted), which is a one way thing for passive sonar, two way for active, etc., etc.

The long tail SQR-19 has a listening window to a frequency range of 0 to 1800 Hz. Ships towing the SQR 19 were limited to speeds of 25 knots cause the breaking strength of the towing cable was 27 knots. Lower frequencies were the best to use to classify a target, but the faster you went the fewer lower frequencies you were going to hear. 5 kts was the slowest speed that could be used. We would compromise frequency coverage for area covered by towing at 12 knots. It also helped fuel consumption. Speed and length of tow cable also affected whether or not you got your tail below the thermocline layer. We would normally slow to dip the tail below the layer. We usually towed the tail 1,000 yards back and just above the layer. But it all depends on the sonar range predictions our computers gave us.

Here are some of the usual range Vs frequency data we would get, SQR 19 short tail 300Hz frequency:
0 to 10 kts 1k to 256k yards
11 to 20 kts 1k to 128k yards
21 to 25 kts 1k to 64k yards

There is no easy answer for detection ranges. Sometimes the acoustic conditions can give you ranges literally in the hundreds of miles, sometimes in the tens of yards. The factors that should be modeled in the game I think are the source levels of the various classes of submarines, and of course that depends to a large extent on speed, their relative target strengths, and the environment. In other words, detection should be based on the characteristics of the target and the environment, and to a lesser extent on the characteristics of the sensor.

Different sensors exploit different types of sound. Narrowband sensors use spectrum analysis techniques to display specific sets of frequencies in the lower frequency spectrum (roughly 0-2000 Hz) generated by a target submarine: propellor and prop shaft rotation rates, reduction gears, steam turbines, turbine generators, various pumps and motors, hydraulic systems, hull resonance's, auxiliary diesel engines, main propulsion motors, etc. Each type or class of submarine has it's own specific acoustic signature, and sometimes a unique signature to an individual hull number within a class of submarines. In general, nuclear submarines are easier to detect and classify passively than diesel submarines on battery. Broadband passive sonar is listening to a wide frequency range, and is basically listening to the total noise output of a target (i.e., all of the narrowband components and the general noise created by the hull moving through the water and propellor cavitation). Since most broadband passive sonar is actually hull mounted active sonar used in a passive mode, it is hearing sound generated in the range that the hydrophones are designed to listen to, that is, the frequency of the active sonar... somewhere between 2 and 15 kHz for tactical sonar systems. As a result, broadband passive sonar is relatively short ranged because the broadband noise in that frequency range is rapidly attenuated.

We break Russian submarines down by their main propulsion systems into types. Type 1 Nuke would be the first generation subs.. Hotel, Echo, November. Type 2 is the Victor and Charlie, Type 3 Yankee and Delta (Type 3 is two type 2 power plants), Type 4 was the Papa... one of a kind SSGN. Type 5 was Alfa. Type 6 is Sierra, Oscar, Akula. Type 7 is the Typhoon. Chances are your going to first classify to propulsion type, then refine it further to class. Signatures between subs in a class differ, so given enough time, or a change in operating mode of the sub, it may be possible to classify to a specific hull number. And the variances are due to the fact that there are minor machinery configuration changes between submarines. In some cases we may not get any further than propulsion type... and in some cases not even that. So, the longer you have contact, the more refined your classification becomes over time. Of course, in a wartime situation, it really doesn't matter. If you're in as ASW free zone (meaning you know the exact locations of friendly subs), you're going to attack first and ask questions later....

Noisy Nuke - 1st Gen US (Nautilus), 1st Gen Soviet (Echo, November, Hotel), Alfa, Chinese Nukes, early Victors, Yankees, Charlie, Delta
Older Diesels - Foxtrot, Tango, Juliett, WW2 era US Guppy's
Average Nuke - Permit, Sturgeon, early Los Angeles, later Victors and
Delta's, Typhoon, Brit and French nukes
Quiet diesels - most NATO diesels, Kilo, Japanese diesels
Quiet Nuke - Later Los Angeles, Ohio, Sierra, Akula
Super Quiet Diesel - Upholder, Collins, later Type 209's, AIP prototypes
Super Quiet Nuke - Seawolf, Virginia, possible newcon Russian nukes?

Passive sonar have the longer range than active sonar even if you don't consider convergence zones. With active sonar, your receiver has to catch the emitters signal, which gets degraded based on the distance to the reflector. The degradation, is an exponential function, so, e.g. four times the emitter power results in only two times the range (just as an example). Passive systems, OTOH, are designed to receive very low power sound signals. Additionally, underwater sound can travel in certain sound channels, hence convergence zones, which increases detection ranges. Another difference is the spectrum of the signals you are receiving. The lower the main frequencies are, the longer the range. Whales, for example, use relatively low frequencies for their communications, enabling them to cover hundreds (some biologists claim thousands) of miles for their communications.

You should also look into experiments, that are done of the Hawaiian coast, where the US is experimenting with low frequency active detection systems for submarines, kind of an active SOSUS It's very interesting stuff to read. You might find a starter on that in the sci.military.naval newsgroups.

Sonar is a very complicated subject. Sonar ranges can change rapidly during a 4 hour watch. And so many things can affect your performance. Even the alertness of your operator play a big part in the equations used. A person on watch for 6 hours is a lot less likely to spot a target that someone on for 1 hour. Sonar duplication in the biggest challenge for a designer of games.

Active Sonar

The active sonar signature of submarines is based on size, aspect in relation to the active sonar, and whether or not they are anechoic coated, and to a lesser extent, hull materials used (i.e. steel, aluminum, glass reinforced plastic). Target strength of a surface ship is a function of draft and length, not how quiet they are against a passive sensor.

A realistic rating would be to base target strength and resulting ranges on the size of the submarine, i.e., SSBN's having the largest target strengths, small coastal diesel submarines the smallest. Assume all later generation Russian SS's/SSN's (from Kilo and Victor III on) have anechoic coating. Anechoic coating is also on all US Flight II and III 688s, Seawolf, and will no doubt be used on the Virginia class as well. Not sure about other submarines. Using a beam aspect US Flight II 688 (anechoic coated) as a target example and assuming ideal acoustic conditions (isovelocity conditions, no sonic layer, in deep water providing the longest direct path ranges) vs. the following sensors:

BQQ-5: 15 - 20 nm LF
SQS-53A: 12 - 15 nm LF
SQS-53C: 15 - 20nm LF
SQS-56: 8 - 10nm MF
SSQ-62 DICASS: 3 - 4nm HF
AQS-13F: 6 - 8 nm HF
Mk46 torpedo: 1000 Yards VHF

The increase in range between SQS-53A and SQS-53C is a result of improved signal processing and a wider range of operating modes, but the output power of the transducer is identical. These are approximations based on best case acoustic conditions. Temperature gradient, ambient noise, scattering layers, target aspect, etc., will generally result in shorter ranges.

Bow or stern aspect reduces target strength considerably. Target speed plays a factor as well. Most modern sonar systems can distinguish target Doppler, so a submarine moving at speed even bow on to the sonar has a good chance of being detected at close to the maximum predicted range of the day. A submarine moving very slowly or dead in the water is another story. In that case the detection range is drastically reduced... say to 25 % of the maximum predicted range. There are many variables involved...

Active hull mounted sonar produces between 100 to 250 dB We are talking about 250-500 thousand watts of sound energy. There are separate motor generators and huge capacitors needed to develop the kind of energy needed for the SQS-53 output. The water inside the sonar dome does reach high temperatures, but the dome is pressurized, and the water doesn't boil. The heat energy is attenuated somewhat by that, and I have never seen water boiling around the bow of a ship when it is pinging. If the electrical power was drawn directly from the ship's power supply it would dim all lights on the ship every time the sonar sent out a pulse. Needless to say, an active pulse is going to kill any porpoises swimming around the bow of the ship (not something environmentalists and sea mammal lovers like hearing). Active sonar can routinely be heard through the hull of an aircraft carrier four decks above the water line, and the active ship was more than 5 miles away...

Distances achieved against a Foxtrot Sub on batteries in active sonar mode, on good days, normal sea states, 10 to 20 kts of wind :

 Min Max SQS-53A 0 yds 3000yds SQS-53B/C 0 yds 18,000yds SQR-19 A(short tail, 140hz to 1800hz) 0 yds 15,000yds

Passive sonar is really, really short range when a sub is in battery mode. We worked a Juliet for about one week and learned that the ranges were much less, about 50% shorter in active. The max passive range I remember for the Juliet and the Foxtrot while they were on diesel power on the surface was 3,000 yds. But this was with the SQR-19 short tail. Diesel engines give off more low frequencies that the short tail cannot hear, but the long tail can at 5 to 10 kt speeds.

The SQS-53B/C sonar also had two extra modes of operation when active due to their increases power and transducer sensitivity. If the ship was in an area of the ocean where the bottom was hard and not covered with primordial oozy mud then the sound energy could be directed downward in the Bottom Bounce mode to bounce the sound off the bottom like a ball on a pool table. This would better the ranges achieved in the Direct Path mode. If we were in water deep enough (over 1,000 fathoms) the Convergence Zone Mode can be used. The sound energy would be depressed 5 degrees from being parallel with the ocean surface and max energy would be sent. The speed of the sound and mass of the water at deep depths would bend the sound upward from its downward path back toward the surface. If anything was in the sound path some energy returned to the ship as a target. The sound energy would often proceed out to the second CZ zone. One CZ = 30 to 35 miles, Two CZ = 60 to 70 miles from the ship. SQS-53A's couldn't do this very well, and SQS-56's only dreamed of such power.

The SQS-53 is a development of the SQS-26. The major difference was replacing the steel sonar dome of the 26, with a rubber one for the 53, along with processor improvements. The rubber dome improved ranges by about 20 percent. The USS America was fitted with an SQS-26 during the 70's, but it was not a very successful arrangement. If the comment is in regard to why an ASW guy is on a carrier, AW's have been on carriers for many, many years. AW's staff the ASW Module, which is an acoustic analysis center and ASW command and control facility, responsible for briefing and debriefing the helo and (no longer) S-3B crews, controlling them while in flight, and doing post-flight analysis of the acoustic data brought back by the S-3 crews. When the DESRON is embarked on the carrier, the ASWMOD becomes Bravo Xray, and is responsible for controlling the entire ASW battle within the battlegroup, and is usually also responsible for waterspace management... i.e., controlling the submarines assigned to the battlegroup.

The US does not employ VDS Sonar any longer. It was phased out probably 10-15 years ago, along with the demise of the Knox class frigates. One of the problems inherent with hull mounted active sonar is it's inability to detect submarines reliably below "the layer." Depending on the time of year, the geographic location, and the weather, the water temperature can remain relatively constant down to a given depth, say on average 200-400 feet. Below that, the temperature declines rapidly (the thermocline), until you reach the region of "deep cold water," which remains constant, a couple of degrees above freezing. Temperature is the biggest single factor affecting sound speed. Without getting too technical, temperature and pressure changes affect the way sound rays travel from a source. Basically, the layer acts as a semi-permeable wall. Depending on the angle at which a sound ray approaches the layer, some of the sound bends or reflects off the layer back towards the surface, while the sharper angle sound rays penetrate through the layer, and change direction significantly. Basically, this causes a sonar hole, or shadow zone below the layer which is a great place for a submarine to hide. By the way, none of this is classified. You can read about this in any good college textbook on oceanography. The purpose of VDS sonar is to lower an active sonar transducer below the layer and eliminate the shadow zone. The downside to VDS is that it is a relatively high frequency, low power sonar and those characteristics are driven by the size of the transducer which has to be small. High freq and low power means relatively short detection ranges. Another drawback is that it seriously hampers ship maneuvering.

What sonar have the longest range, Active or Passive? There is no simple answer. Again, it depends on the target, how advanced the sonar system is, and the acoustic conditions. From the submarine's point of view against surface targets, passive sonar is the longer range sensor. Ships active sonar is the sensor of choice against a diesel sub. However, active sonar has to be used sparingly and with good judgment. Normally, we would use other sensors to at least get an idea of where the sub is. Third World countries, by the nature of their command and control systems, generally communicate frequently with their forces at sea, so HFDF and other SIGINT methods are used to narrow down an enemy subs position. Then aircraft are generally vectored into the area using radar to look for periscopes. As a general rule of thumb, active sonar counter-detection range is three times the detection range. When you light off an active sonar, you just broadcast to the enemy where you are (just as with radar), and an aggressive sub commander is going to come towards you, not be scared away.

Assume active sonar can positively ID a contact as a submarine and narrowband passive sonar (passive sonobuoys, towed arrays, and SOSUS), can positively ID to class. Of course this is simplifying things a bit... the ability of a passive narrowband system to ID a contact depends on a number of factors... target speed, range to the sensor, system sensitivity, operator proficiency, to name a few... and in general, it is easier to passively classify a nuclear submarine than a diesel operating on battery. And on one other note, something I can't get into much detail about... there is fundamental difference between US/NATO nuclear submarines and Russian nuclear submarines electrical systems, so it is easy for an experienced acoustic analyst to quickly tell the difference between the two, even if you don't otherwise have enough of a signature to identify to class.

Soviet active systems were very compatible to USN systems. They had a lot of power to make up for their weakness in the area of signal processing. In some cases two Soviet subs in the Mediterranean exchanging data via a sonar link when they were 100 miles apart. In the passive system arena the west was far ahead of them because of computing power. For example, the original SQR 19 towed array fed its data into multi-linked computers for processing.

For many years, active sonar has had the ability to define the shape of a target. The SQS-53, AQS-13 dipping sonar, and the VLAA system have that capability. However, they do not have the resolution to actually classify a submarine as to type. That requires very high freq sonar, which are extremely range limited.

As a general rule of thumb, active sonar can be counter-detected at least 3 times the predicted range of the day.

Dipping sonar

Dipping sonar were always a high frequency sonar. It couldn't be low freq because low freq sonar must have large transducers to catch the large low frequency sound waves.(Same kind of relationship that radio has with its antenna sizes.) The Soviets had a good dipping sonar because they had a big helo flying the gear around.

Sonobouys

The current inventory of US sonobouys includes the following:

SSQ-53 DIFAR (Directional LOFAR)
SSQ-57 LOFAR (Low Frequency Analysis and Recording - omnidirectional)
SSQ-62 DICASS (Directional Command Activated Sonobouys System - Active)
SSQ-77 VLAD (Vertical Line Array DIFAR)
SSQ-110 VLAA (Very Low Frequency Active Acoustic)

The SSQ-110 is a buoy that deploys a line charge that can be remotely detonated to generate a very low freq active pulse that propagates a long distance. Standard SSQ-53 DIFAR buoys are used as the receivers. This is the old Julie system brought up to date, using advanced signal processing, and is said to be very effective.

There are other experimental and special purpose buoys (i.e., submarine communication uplink and downlink buoys, etc.) that are not in general use in the fleet. As for detection ranges, they are target and environment dependent. Sonobouys are not as sensitive as some other underwater sensors, mainly because of the size of the hydrophones which are very small. They are not designed to be long range sensors anyway. DIFAR buoys are less sensitive than LOFAR buoys because the bearing signal is uplinked to the aircraft over the same carrier as the acoustic signal. NATO uses US, British, and Canadian manufactured buoys. The designations of the British and Canadian sonobouys are different, but they have the same characteristics and capabilities as US buoys. Typical ranges:

-LOFAR vs. diesel/electric on batteries = less than 1000 yards on a great day. Even when you know it is there.
-LOFAR Vs diesel/electric snorkeling or surface = 3,000 to 12,000 yards depending on the sea state. Against noisier subs and surface ships they could hear a long way. Better than a towed array due to the lack of self noise. But the towed array was better at identifying the target cause it could listen down to 0 Hz. If the water is over 1,000 fathoms deep and around sea state 2 or less, CZ returns from LOFARs is possible. P-3's often reported the merchant traffic in the area after they laid their buoy patterns because they heard so much.
-DIFAR gives identical distances as LOFAR but with a relative bearing to the noise.
-DICASS was also the same ranges in passive mode. But in active mode it was a high freq sonar without a lot of power so a max range of 5 miles when active might be pushing it. The DICASS active mode has shorter range than passive mode. Passive buoys have always been better than active because they can hear low, Med, and high freqs but can't send out anything but high frequencies.
- CASS Buoys have been retired. Basically an early version of DICASS.

As far as sonobuoy loads, US aircraft carry the following maximum loads:

P-3C: 84
S-3B: 60 (Note: ASW mission has been taken away from S-3 as of Jan 99)
SH-60B: 25
SH-60F: 12
SH-3H: 12
SH-2F/G: 15

Bouy loadouts depend on the anticipated target, and in the case of the S-3, launcher slots are also used to deploy chaff and flares, so normally did not carry a full load of sonobouys. Against nuclear targets primarily passive bouys are carried, and versus diesels primarily active bouys are carried. I also forgot to mention another bouy type in my previous e-mail, the SSQ-36, which is a bathythermograph bouy. A P-3 will normally carry 60 passive bouys and the remainder active and BT bouys.

The Sea King has 12 tubes, but the tubes can be reloaded from inside the aircraft. It can carry as many as 30 extra buoys inside the aircraft on ASW ops. The total number can be increased but this would have to be figured into the weight and fuel load. P-3's can, and occasionally do, carry more than the normal load of bouys. On SH-2 it was not uncommon throwing buoys out the door.

Anechoic coatings

Anechoic coatings were designed mainly as a countermeasure to defeat acoustic homing torpedoes. Essentially, they absorb the very high frequency active pulses of an active homing torpedo, reducing the acquisition range substantially... something on the order of 30 percent. Against high power low freq sonar, the coating is less effective, say maybe a 5-10 percent reduction in detection range.

Prairie and Masker come into play only for passive sonar searches. They make the ship sound like a rain storm to the submarine. The sub knows something is making noise but can not identify the noise as any type of a ship. Prairie and Masker play no part in Active Sonar use. If you are banging away with 235dB of sound energy the sub will know where you are.

Prairie Masker is very effective. Even our submarines have a difficult time locating our CG's, DDG's DD's, and FFG's when they are operating in condition 2AS (ASW stations) with prairie masker active during exercises.

A carrier operating with an effective acoustic deception plan can still be detected at long ranges, but can be mistaken for a different target... i.e. a merchant ship or a smaller combatant. Acoustic deception plans have been used it in exercises very effectively.

Has anyone heard about the Mini Mobile target simulator? They are deployed for tracking practice. What would happen if the sound tape that they carried was replaced with one that sounds like a carrier? Force a sub down deep, deploy the simulator, wait for him to shoot at it and then put your torp right on top of him.

Sonar vs. torpedoes

A torpedo, regardless of type (and assuming there are sensors listening for it) will be detected very shortly after launch. In the case of the rocket torps, which have a speed of 200-300 knots, and are strictly nuclear tipped, it probably doesn't matter. These rocket torps have no internal guidance system... just point and shoot and on a timer. The sonar operators will maybe have enough time to alert the tactical action officer and then bend over and kiss their asses good-bye...

Torpedoes are noisy, very noisy. The propulsion type of the torpedo (thermal, steam, electric) would have the greatest effect on noise, but that is relatively insignificant since they all end up spinning propellors at very high RPM (except those nasty rocket torps the Russians have). Under average acoustic conditions, torpedoes are going to be detected as soon as they are launched, taking into consideration the time it takes for sound to travel in water. Sound speed in water is 4800 feet per second at the surface, water temp 39 deg F, salinity 35 parts per thousand. Assuming the water is warmer and the torp is launched at a depth greater than the surface, sound speed is going to be faster, say around 5000 feet per second: Roughly a mile. If the sub launches the torp from 20 miles, the receiving sensor will detect it approx. 20-25 seconds after launch. By the way, that is a very long range to launch a torp from. Even though some torps have a very long range, tactical doctrine and the realities of having to visually ID the target mean that most torp launches occur inside 7 NM during daytime, and as low as 4 NM at night or in bad weather. Sub Vs Sub torp shots are also going to occur at close quarters for the most part. So to simplify that aspect, assume torps are automatically detected inside 20nm. In this case, torpedo size doesn't matter...

Regarding the Type 65 wake homer torp... The range is somewhere around 50 miles. However, Russian doctrine calls for torpedo launch at no greater than 7nm. Remember, torp is going to be heard shortly after launch. If it's launched 50 miles away you'd have time to outrun it. And the warhead is close to 1000kg... or approx. 2200 pounds. Pretty confident that one exploding under the keel would break a carrier in half.

Submarine operations

How can you be sure that the sub you are firing at is an enemy sub? To bad if you sink one of your own \$1 billion subs.

You can't be sure, at least not absolutely. OTOH, subs are normally assigned certain patrol areas. Any submarine that they detect within that area then has to be considered hostile since no other of your own subs will be permitted into that patrol area. There are exceptions but then there will be a "safe passage" arrangement. Normally, submarine patrol zones in a declared war zone are "kill areas". Disclaimer: AFAIK, this applies to SSK tactics in brown-water areas. SSN tactics might be different especially if they are protecting boomers.

There is a standard NATO procedure for establishing patrol areas, safe areas, etc., which we call Waterspace Management/Prevention Of Mutual Interference. This serves two purposes. One is preventing potential collisions between friendly submarines, the other is to avoid a blue-on-blue engagement. Think of it as air traffic control for submarines. Same procedures apply whether it is brown water ops or blue water ops. A pretty complicated thing. Another way to be reasonably sure is when and if you have good passive detection of a submarine and have had time to analyze the acoustic signature. In the 70's and 80's we had developed enough acoustic intelligence on Russian subs to be "acoustic positive" to a specific class or variant of Soviet subs, and in some cases down to an individual hull number.

 -