2018/02/18

Modern warships (V) - land attack

.
I don't care about naval land attack capabilities.

Land attack by sea is not about deterrence or defence. 

NATO powers could use naval cruise missiles, but hardly anything of interest should be farther than 500 km from friendly territories, and thus there's no reason why we couldn't use land-based missiles instead. If need be, we could use air-launched cruise missiles, even dropped from transport aircraft.

Cruise missile diplomacy is a violation of the Charter of the United Nations article 1,  the North Atlantic Treaty article 1, Briand-Kellogg Pact article 1 and Revised General Act for the Pacific Settlement of International Disputes that are in effect and thus have the force of law in all countries that ratified them (or had their legal predecessor ratify them).
Small wars are a waste of resources and almost all of them are also violations of the aforementioned international treaties.

Substantial naval cruise missile land attack capabilities are important only to despicable acts; cruise missile diplomacy, offensive small wars and strategic surprise attacks.

Naval cruise and (conventional) ballistic missiles of greater than 500 km range should be banned in my opinion. There's no legitimate justification for their existence. Nuclear-tipped ICBMs and SLBMs should be handled differently, as part of nuclear disarmament or move towards minimal deterrence regimes.

The only land attack mission of a navy that may be worthwhile and legitimate is to raid pirate havens, for that's how competent navies deal with pirates. See Pompey the Great. Pompey and his fleet wiped piracy off the Med in weeks (after months of preparations), primarily by going after their bases. The current crop of navies pretends that patrolling against pirates (=job creation scheme) is the way to go. No, it's not. You do intelligence, then you raid the pirate haven, blow up all boats, blow up the leader's villa and return home. This requires no more than some infantry (whether marines or regular infantry doesn't matter), some offboard motor-driven RIBHs (rigid hull "inflatable" boats)  and a small chartered cargo ship.

A reconquest of islands occupied by an aggressor should be avoidable by using embargos and blockades against the aggressor instead.

No warship needs to be set up for land attack.

A little land attack capability may be for free as AShM and guns may shoot at land targets as well, but that should have no priority.

S O
.

2018/02/16

Modern warships (IV) - ASuW

.
Land-based and carrier aircraft can attack naval surface targets much more easily, at much less risk and all this while being able to identify targets at a longer distance than surface warships themselves can do. Fast attack craft with missiles are thus an anachronism, and anti-surface warfare capabilities have become an afterthought for the design of warships.

Land-based strike fighters could reach a ship anywhere in the North Atlantic
with an anti-ship missile if supported by tanker aircraft.

The last naval warfare campaign in which surface craft were important was the naval blockade of the Sri Lankan government against the Tamil rebels. They used mere patrol boats to intercept blockade-running boats used by the rebels to import supplies from nearby India.

Other than this the unimportant Battle of Latakia in 1973 between Syrian and Israeli missile-armed fast attack crafts was the most recent relevant sea battle. Air power could easily have substituted for either party, but the air forces were busy apparently.

The first serious people understood in the First World War that air power could wipe navies from the surface of the sea within its effective range. Some bombs used and the first aircraft-dropped torpedoes had been developed, and even some guided weapons had been tested.

300 kg guided anti-ship glider, to be dropped from airships (1917)
Aircraft can synchronise attacks from multiple preferred angles with a well-timed application of anti-radar practically any anti-ship missile type (all of them are or could be adapted for air/sea use). They can do so while being a very fast and difficult-to-hit target, particularly at distances greater than about 40 nm. They can also provide standoff jamming and chase away or destroy AEW support.

Scenarios

There are few legitimate scenarios in which frigates and destroyers might need to do ASuW in absence of any air support other than naval helicopters. Some of those are:
  • Passing through a strait and encountering Q ships or small boats / wing in ground effect craft
  • Engaging an auxiliary cruiser on an ocean after being attack with missiles, possibly from its helicopter(s)
  • Ship battle after failure of either side's air power to sink the warships
  • Surprise sea battle at the beginning of a conflict
  • A sea battle including at least one poorly equipped navy (imagine a Western navy would escort humanitarian transports to a Biafra-like conflict zones and getting engaged by a desperate Third World navy that wouldn't be destroyed before it opened fire itself)
  
Non-scenarios 

Still, I don't think the Taiwanese navy should for example equip warships to deter or sink a Chinese invasion fleet. It would be much more cost-efficient and thus much more effective to invest in land-based missile batteries instead. (Of course, their navy likes having toys at sea and thus they even have utterly pointless fast attack craft).

The extremely fashionable "Iranian speedboat threat" hype that appeared in 2002 during van Riper's use of simulated speedboats is ridiculous in my opinion. You won't have any trouble with Iranian speedboats if you don't attack Iran, and any powerful hostile country would be capable of much worse. The security of Kuwaiti and Saudi oil exports in the Persian Gulf is their problem. They can invest in a pipeline or two to the Red Sea for a few billion dollars. There's no reason why Western navies should prepare for war against Iran, and hardly any other scenario than another Gulf War/Blockade attempt for going that close to hostile shores in anything but a war of aggression. The last time the West escorted tankers against Iran was in support of Iraq waging its war of aggression against Iran - and Iranian oil exports were not protected, so it was at best a hypocritical Western campaign in support of an aggression.
As a rule of thumb NATO forces do not need to get close enough to hostile shores to fear speedboats in a defensive war. Such things might only happen in wars of aggression or stupid small wars.

Gap filler

ASuW by frigates and destroyers is thus nowadays a mere gap filler for when air power is not available. It was accordingly neglected by Western navies, which are still operating their 1970's generation of anti-ship missiles while the Russians kept innovating and refining their arsenal.
The typical armament of a Western warship for ASuW consists of 4 Exocets, 8 Harpoons or rarely 16 Harpoons and also a single or at most two guns of 57...127 mm calibre. There are a few other Western ship-launched missile types, especially Otomat (Italy) and RBS-15 (Sweden). The Otomat with its launcher is a bulky design similar to some Russian AShM launchers and to find or free up deck and roof areas for such launchers is difficult.
Guns of 57...127 mm calibres are not really ship killers, but they can disable and set afire ships with enough hits at the right places. The 57 and 76 mm calibres are rather boat killers than serious against ship targets, though.

The gap filler role of ASuW for frigates and destroyers means that relatively little money, volume, deck/roof area and weight should be allocated for ASuW. It's also quite important that weights mounted high on a ship are much more troublesome than weights located rather low - this is about top-heaviness (rolling) of ships and about metacentric height.




Some navies appear interested in joining the ASuW and land attack mission into one type of missile and one type of gun (the latter usually 127...155 mm), which would make some sense if land attack as a mission made sense. It doesn't make much or any sense in deterrence and defence, but it's a favourite child of great power gaming fans.

Basic tactics

There are formulas from operational research that are too far from realism to be of relevance - the Lanchester equations, for example. The insight that those who shoot first may hit before the others launch their missiles helped the Syrians very little in 1973. Their initial salvo was an utter failure, the Israeli return salvo a success.

To avoid being surprised seems to be the most important rule for naval combat if not combat in general.

Modern ship-to-ship combat may actually place a 100% emphasis on dealing with the naval helicopters. These can carry AShM, but more importantly they are the forward eye in the sky for their surface allies. To take out these helicopters could thus be the most important part of a modern sea/sea battle. Long-range area air defences would be expected to take them out if neither naval nor land-based fighters were available. That's a point in favour of carrying a few SM-6, for its effective range may be about 250 nm against helicopters according to published info.

A towed, tethered aerial sensor (project TALONS or a similar autogyro approach) could have a much farther horizon at altitude than mast-mounted sensors, but this would still not be a replacement for a helicopter. Moreover, I expect such tethered systems to be used at a much lower altitude for a simple reason; targets can be detected best on the horizon, when there's no sea in the background. Most sensors that could be lifted to 400+ m would have a much smaller effective range than the horizon distance at such  altitudes (almost 50 nm at 450 m altitude). Infrared sensors that do not betray the ship's location with radio emissions might have their effective range against missiles matched with the horizon at about 50 m altitude already, for example. 450 m towed altitude might be useful for ESM (passive direction finder for radio/radar emissions), though.
 

Another very important issue is to detect and identify early, and to make decisions without unduly long lags. This means especially that the tactical commander has to have sufficient confidence in a positive identification of a target. He couldn't simply unleash firepower of great destructiveness on what might be cruise ships or fishing boats. Even an utterly amoral commander wouldn't want to do so because wasting munitions on false targets might be the doom of friendly forces later. You got to get a positive identification of high confidence.
Anti-ship missiles with two-way HF datalink and imaging sensors might help with this if their radio connection wasn't cut, and relatively cheap drones with HF datalink could help a lot as well. They would either be shot down/jammed (confirming hostile presence) or report back.
No such drone is known to me as being in naval service. There were some projects, but none competed successfully with naval helicopters. Too bad killing those helicopters might be rule #1 in sea battles. In the end, sea battles are so unlikely that hardly any modern surface warship is really well-equipped for one. The Russians have some heavily armed (for ship killing) cruisers and destroyers, and that's about the most extreme there is.

Long range anti-ship fires offer a particular problem; the subsonic missiles may travel for 300 nm or even a thousand nm if built for it, but the targets would have moved by several nm and a convoy may have changed its formation. Targeting data input from before launch would often be obsolete, leaving the missile's computer all alone in its attempt to figure out which contact to engage. It might even hit some neutral ship that's in the wrong place at the wrong time. The answer to this is straightforward; a datalink by radio. The missile could at very least receive targeting updates, if not even send back processed sensor data. Datalinks aren't necessarily reliable even in peacetime and whatever eye in the sky was tasked to provide data for updates might be gone by the time the missile needs the updates. A higher average speed (supersonic missiles) and a limitation to shorter engagement distances can be used to make do without targeting updates by datalink. Alternatively, one might put neutral ships at risk and simply launch more missiles to make up for those that hit decoys and already wrecked or otherwise unimportant ships.

The supposed speedboat threat is similar to the heavyweight torpedo threat in that one might try to run away at 30+ kts even though the speedboats are faster. This buys more time for countermeasures, though mostly fast ships (running away wouldn't help 15...25 kts cargo ships as much). The installation of the gun on the forecastle of the LCS classes was thus plain stupid and regarding ASuW at best explainable with aesthetics. (The forecastle location makes a bit more sense for CIWS purposes). A ship that runs away from speedboats would want to be able to fire to the rear 30°, or else the last several knots of its top speed that were purchased at great expense of other ship characteristics would be wasted. The ship would not be able to move straight away from the threat and shoot at it with a forecastle gun at the same time.

What does it take to sink a ship?

Chuck covered this in several blog posts, so I'll simply link to him:
https://chuckhillscgblog.net/2011/03/14/what-does-it-take-to-sink-a-ship/
https://chuckhillscgblog.net/2012/04/06/what-does-it-take-to-sink-a-ship-illustrated/
https://chuckhillscgblog.net/2014/10/02/what-does-it-take-to-sink-a-ship-another-illustration/
https://chuckhillscgblog.net/2016/09/16/what-does-it-take-to-sink-a-ship-the-ex-uss-rentz-sinkex/

a heavyweight torpedo hit on a rather small warship
In the end, shells and missile warheads damage ships very much and often set them on fire, while torpedoes go to work opening the hull to seawater with often much more decisive effect regarding the question of whether a ship is floating or not.

ASuW with missiles

Area air defence missiles can be used to damage warships, and surface-to-air missiles have indeed damaged warships in accidents already. A firepower kill is feasible even with the small fragmentation warheads of such missiles, and a salvo of SM-6 (64 kg warhead, very long range against ships, up to Mach 3.5) missiles could weaken the defences of a targeted warship to such a degree that obsolete Harpoon follow-on missiles could score hits even without the element of surprise.
Some modern air defence missiles (with CIWS/short range and area air defence types) have demonstrated the ability to destroy tiny boats, even within well less than a nm distance. That's a nice to have backup, but warships shouldn't really come into contact with speedboat threats anyway.

Some AShM types are claimed to be able to fly evasive manoeuvres in the terminal approach to the target to improve survivability especially against gunfire, but it's unknown to me if any missile does so in response to actual targeting (or just as a matter of autopilot behaviour) and it's unclear just how effective this is. I suppose it would reduce the distance at which gunfire can intercept the missile, and thus increases the chance of missile wreckage or explosion fragments hitting the ship target.


There are no publicly known dedicated anti-radar missiles in shipboard use nor any publicly known anti-ship missiles with ARMs as submunitions. ship-to-ship missile combat thus lacks the ARM element that air/sea combat may have.

Naval helicopters can launch anti-ship missiles, but they don't have much payload, so the heavy missile types are de facto unavailable. Two lightweight torpedoes is a typical payload for such a naval helicopter - that's twice about 250-300 kg. Almost all AShM weigh more than 400 kg, save for rare lightweights such as the Marte. NSM is rather light at 410 kg. The typical Western anti-ship missiles  such as Exocet and Harpoon are rather associated with medium and heavy helicopters, while the lighter end of medium helicopters tends to employ lightweight missiles such as Marte, Sea Skua, Penguin. The latter ones were still highly regarded as missiles against fast attack crafts (boats smaller than 400 tons), as those lack area air defences. To launch an anti-ship missile from the sensor platform itself is less of a technical challenge than a networked engagement (helicopter detecting, ship firing a missile).


Helicopters can switch between being in the radar horizon and being below it quicker than area air defence missiles with semi-active radar homing can exploit this for a kill. This allows for a quick radar scan and then hiding again, just as the Super Étendard pilot did who sank the HMS Sheffield with a single AM39 Exocet missile.

Anti-ship missiles of all categories mentioned in the AAW article may be launched by warships even though some missile types have no version for shipboard use (Sea Eagle, Kormoran and Kormoran 2 had none, for example).

ASuW with missiles is of course very much under influence of what was written in the part about AAW, as  a sea battle between capable warships would be about attack and defence.

ASuW with guns
 
Some post-WW2 warships were built without any guns. They didn't seem to need any, but this was later corrected and at least one gun (57 mm or bigger) is accepted as minimum backup to missiles today. The Falklands War and especially the air attack as San Carlos bay helped to dispel the reputation of in particular the British air defence missiles which were much less useful than their public reputation up to the Falklands War. Short range defences gained a lot of attention due to the Falklands War.

The 105 mm calibre was ridiculed as being of little value in ship-to-ship combat as early as the beginning of the 20th century, even as a deterrent or defence against the torpedo boats then in use. The explosive power of such a shell is too small. Modern 100 mm shells pack a better punch and proximity fusing makes it easier to achieve some effect on small targets (shell exploding above a boat and showering it with fragments), but the fundamental problem remains. Even 127 mm shells are of little sinking power unless they hit at the waterline. Submarines of the world wars sank scores of ships with 88 mm and bigger guns by getting very close and piercing the ship hull at the waterline. A blunt instead of rounded shell nose may actually help with penetrating the water on the final meter without ricochetting. Still, there's little ship-sinking potential in all naval gun in use world-wide unless the target is already utterly defenceless and needs nothing more than scuttling.

This fits to widespread opinions from the 20's to 40's which saw the lower limit for effective ship-to-ship combat guns at 150...155 mm, possibly 139 mm (the French view). 139 mm may thus be the lower limit of satisfactory ship-killing power with modern shell technology (better steel for thinner shell walls, more volume for better explosives). Attempts to install new guns heavier than 130 mm calibre in post-1960's warships were not really successful. Neither the navies nor the developers were able to resist automatic loading, which makes the whole thing much more expensive and heavy than necessary. One could employ a manually loaded 140 mm naval gun from the 1910's and mate it with a 100 or 127 mm turret's automated elevation and traverse control and would end up having the best if not only truly ship-killing (and naval gunfire fire support) naval gun system in service world-wide. Nobody seems to see a need for this, though. The last Western cruiser class with really good ship-killing gunfire capability was the Tiger class. The Zumwalt class "destroyers" have 155 mm guns, but those have munitions issues and are meant for land attack.

The typical guns on warships are thus of 76, 100, 127 and 130 mm calibre or smaller. A well-proportioned gun looks really good on a forecastle, but two 76 mm guns (mostly for AAW) or a single aft and low-mounted 127 mm gun seem to be the rather sensible choices. The latter would have little air defence capability (especially the American 127mm turrets that are lighter but halve a much lower rate of fire than Italian 127 mm turret designs), so a separate CIWS for the AAW role would be advisable.

76 mm L/62 Super Rapid turret (120 rpm) without munitions (and without optional STRALES): 7.9 t weight
127 mm L/54 Mk 45 Mod 2 turret (16-20 rpm) without munitions: 22.2 t weight 
127 mm L/64 turret (33 rpm) without munitions: 33 t weight

ASuW with torpedoes

Torpedoes used to be important in surface actions, but their short range and slowness eliminated them from being serious ASuW munitions for surface warships and boats. Their success depends too much on the element of surprise. Torpedoes could still be of use in scuttling crippled ships, but even lightweight torpedoes could do this. Only the latter can be carried in satisfactory quantities for ASW, so adding heavyweight torpedoes for ASuW and ASW in addition to ASuW missiles and ASW lightweight torpedoes is an unnecessary expense and adds unnecessary weight and manpower/training needs. The torpedo designs are available (the submarines' torpedo designs could be used), but it simply makes no sense to use them on surface warships.

Other remarks

ASuW missiles could be programmed to include an anti-helicopter mode, and be a little cheaper than long range area air defence missiles such as SM-6 in this role.

Helicopters have a theoretical ability to drop bombs and unguided PGMs, but I suppose that might at most be a niche with tiny PGMs or smoke munitions in support of boarding (fast rapelling/fast roping) actions. Doorgunners with machineguns are more relevant for this.

AShMs with two-way datalink could cooperate (fashionably: "as a swarm") and use their different perspectives to triangulate targets with IIR or passive radar, to report identified decoys as well as to synchronise their attacks to locally saturate defences.

Imaging infrared sensors (mounted on missiles) may be able to detect the muzzle flash of guns and the hot gases of air defence missiles. This could inform them about necessary evasive actions.

Imaging infrared sensors may also have the ability to tell already burning or sinking targets (pattern recognition showing the ship is already broken apart et cetera) from unscathed targets, informing the missile's decisionmaking in regard to target selection.

Supersonic missiles will likely not use IIR sensors because the friction at supersonic speed in the dense very low atmosphere would heat up the sensor's window. Good IIR sensor windows would also be aerodynamically inefficient.

Both the UV and the visible spectrum are dependent on daylight and even more weather-dependent than IR, and thus practically irrelevant for AShMs.

Conclusions

Again - as with AAW - an eye in the sky is most important and necessary for beyond/below the horizon fires (unless some other surface unit is more close to the target). Naval helicopters can detect, identify and even engage surface targets up to sea state 5 or 6. Long-range AAW missiles such as SM-6 might be useful in blinding the opposing force by taking out or suppressing their eye in the sky.

Area air defence missiles are relevant in ASuW even in an offensive role, especially to disable the air defence capability of the target. They may also be crucial keeping fixed and rotary wing aircraft from providing a positive target identification and targeting data for missile attacks.

Shipborne cruise missiles used to be installed in dedicated launchers on deck or on superstructures, but AShMs really should be loaded into vertical launch silos. This offers the advantage that opposing forces don't know the quantity of AShM that a warship has and that loadout can (in theory) be changed as estimates regarding ASuW threats and needs change. It also helps with radar echo reductions and generally frees areas and volume up for other purposes. Missiles in a silo are furthermore less of a secondary explosion hazard than relatively exposed missiles protected but by a tube. An effective  hit on an otherwise loaded VLS would be catastrophic even without the presence of AShMs anyway.

The long range of modern AShMs suffices and seems to make the armament of helicopters with AShMs unnecessary at least when a HF datalink to the AShM can be maintained by either the helicopter or the warship itself.

A combination of multiple anti-ship missiles seems unnecessary for a warship albeit it would be a nice to have for air power. A stealthy subsonic sea skimmer with great target identification abilities and the ability to aim at particularly important parts of a ship would be well-justified in moderate quantities. The total reliance on a IR-dependent missile may be inappropriate given the weather dependence of its sensor. A missile with both radar and IIR might be more reliable and adaptable. A mix of two different AShM missile concepts (stealthy IIR seeker sea skimmer and supersonic active+passive radar seeker sea skimmer) might make sense due to its desirable redundancy.

The range of AShMs does not need to be greater than 200 km and at least a datalink for the upload of new waypoints and target locations to the missile would be promising, while a two-way datalink to a helicopter and other AShMs would be even better.

The weight advantage clearly favours two 76 mm guns or even but one 76 mm gun as a warship's gun armament, and this should continue to be the standard for Western frigate designs and possibly warships in general (in my opinion). These guns do little ship-killing in ASuW, but they can cause a mission kill on a nearby warship or boat just as much as heavier calibre guns can do and they are lightweight and relevant as CIWS.

S O
.

2018/02/13

Modern warships (III) - AAW

.
(This is more a like book chapter than a mere blog post; about 8,700 words ~ 20 book pages.)

Intro

Anti-air warfare (AAW) is mostly about protecting ships from threats in the air, rather rarely about helping fighters in an air superiority mission or protecting coastal objects such as ports from aerial threats. (I will write about ballistic missile defence separately.)

Air defences on land have a mixed reputation for their ability to provide area air defences. Combat aircraft regularly entered the defence zones of area air defences and got away with it. This was usually either due to their exploitation of terrain features such as hills and mountains (which aren't available at sea) or due to sophisticated countermeasures against the area air defences, especially jamming and missile attack on active radars.
(Western) Naval area air defences seem to command more respect regarding defence against aircraft, possibly because they had few opportunities to reveal shortcomings (such as in the case of HMS Sheffield). Another reason may be that navies have a self-interest in pretending that their surface warships are properly protected against air strikes; to admit the opposite would put the funding for surface warships at risk.

The typical aerial threat of the Second World War was a manned aircraft, but the first guided munitions were already in use by 1943, and the British air defence of 1944 was in large part occupied with protecting against V-1 cruise missiles. This foreshadowed the later importance of missiles.

The focus of warships' air defences has moved from defeating missile launch platforms (aircraft) to defeating missiles decades ago. It's being assumed that the latter would rarely dare to come into range except at very low altitude. This focus on missiles was mostly about classic anti-ship missiles. The threat posed by anti-radar missiles and swarms of hundreds of cheap & slow drones doesn't seem to have attracted as much attention.

French Rafale strike fighter with an AM39 Exocet ant-ship missile
I will try to give a near-comprehensive overview over modern and very near future naval AAW and offer some conclusions.

Surviving the aerial threats 

The survivability of ships at sea in face of aerial threats can be enhanced by many ways:
  • avoiding being found by opposing forces
  • avoiding being identified
  • shooting at aerial platforms
  • shooting at missiles at long ranges (area air defence)
  • shooting at missiles at short ranges (self defence only)
  • defeating the incoming missiles' sensor and decisionmaking (jamming, concealment and decoys)
  • surviving after being hit
To avoid being found is nice as long as it works, but sooner or later some convoys would be found by hostiles.

To avoid identification means to keep the threat aircraft far enough away that it cannot gain informative infrared imagery and one would also need to jam their radar. Radars can create decent resolution imagery in a synthetic aperture radar (SAR) mode. 1980's tech was capable of creating good enough imagery to tell a cargo ship apart from a frigate, but modern SAR could enable the identification of a ship class. It's technically fairly easy to jam this, though this may require treacherous emissions for extended periods.

examples of SAR imagery
I saw much better resolution imagery for SAR used on landscapes.
To shoot at aerial platforms is de facto all about area air defence and quite similar to shooting at missiles, save for ballistic missiles and other multi Mach fast threats that may require different fusing and warhead concepts than the intercept of ordinary missiles and aircraft. The important difference between an aircraft and a missile target is that the aircraft may have much better countermeasures against your missiles because its much greater value justifies the effort. Anti-ship missiles will rather not have serious radar jamming abilities, but most modern strike fighters do and even obsolete strike fighters can be upgraded with some jamming capability at relatively low costs. On the other hand, radar jamming tech has become so affordable that it's included in expendable 55 mm decoys - which may point at a possible presence of expendable decoys or radar jamming in general in modern or future anti-ship missiles after all.

IIR makes ship recognition much easier if you are close enough
To defeat incoming missiles' sensor and decision-making is typically about onboard radar jammers and projected chaff clouds, but projected multispectral smoke clouds, newer types of decoys (freely moving drone dinghys, towed surface decoys, free flying decoys), heliborne jamming, cooperative jamming, dazzling lasers against IR sensors, changing of ship shapes (as with disguised WW2 auxiliary cruisers) may be used as well. One may also use entire relatively unimportant ships as decoys or even as cover.

Survival after being hit is no more about keeping the warhead outside as it was with battleship armour plating. It's about limiting how much of a ship gets affected by the blast and fragmentation effects of a hit. It's about offering little fodder to fires, extinguishing fires quickly, limiting the amount of water the ship takes in when leaking, keeping onboard components redundant and having the ship's tactical command centre at a relatively safe location (below the waterline). A large size helps as well. Continued operation after a hit depends a lot on the crew and the availability of tools and spares needed in such a situation. A ship that's got no more than the absolute minimum crew size won't cope well after taking casualties even if the hardware stayed functional.

It's important to remember that a warship's purpose in battle is mostly to protect. The defeat of threats is but a means to this end. A perfect defence of the warship itself is of little use if meanwhile all escorted transport ships were destroyed. A warship's self-defence and ability to continue operation after being hit is thus mostly a means to the end of continued protection for the protected ships (transport ships, carrier).

Area air defence

Area air defences are necessary for a proper defence even if close-in weapon systems were near-perfect simply because they're the only defences with an effective ceiling that keeps threat aircraft from safely bombing ships with cheap guided bombs at will.

The most famous naval area air defence system is AEGIS. A convoy with ideal AEGIS defence has a so-called cruiser (CG) that's effectively an AAW command ship and AAW escort in itself and several destroyers (DDG). Large phased array antennas on their superstructures search the sky for threats, and (originally) illuminator radars enable semi-active radar homing missiles to fly towards aerial line of sight targets. The SM-2 missiles need target illumination by radar only in the terminal phase and fly by autopilot on intercept course most of the time. Thus multiple threats can be approached by SM-2 missiles per illuminator radar onboard. Airborne early warning (AEW) may be provided by Hawkeye aircraft or not. The close-in weapon systems were rather unconvincing on early AEGIS ships, but there were many chaff projectors and substantial radar jammers in use.

This late Cold War air defence concept was oriented against massed attacks by Soviet-style anti-ship missiles, or whatever of such attacks made it past F-14 and eventually also F/A-18 naval fighters. Those Soviet missiles were large, mostly supersonic and flew quite high. AEGIS was expected to usually have a line of sight to them and pick up the radar echoes at long ranges.

SM-6 launch. Note the booster stage;
it's an indicator for a long minimum
engagement distance.
Western-style anti-ship missiles such as the famous Exocet (but also Kormoran, Sea Eagle, Harpoon, Gabriel and some others) were small, subsonic and (since the 70's)  built to fly at very few metres altitude (as low as 4 m) to avoid detection until it's too late ("sea skimmer" missiles). AEGIS was poorly suited against this until about 2004 when finally ESSM missiles appeared that are of use against such sea skimmers not only in tests, but potentially also in wartime. SM-6 missiles have with an active radar seeker appeared in the meantime; they further add to the ability to intercept sea skimmers, albeit this is little more than theory. A single SM-6 missile costs more than many anti-ship missiles do. It's rather an anti-platform (mostly anti-aircraft) missile and a limited ballistic missile defence munition than a munition for defence against cruise missiles. It's also questionable whether modern sea skimmer missiles with radar 'stealth' would be detected by the ship at ranges where the SM-6 is without onboard competition as a defence.

Sea skimmers can be detected by a warship at ranges shorter than the radio horizon, and one may squeeze out a few seconds earlier detection with a mix of radars and horizon-scanning IR sensors. Ideally, one would have some picket that detects them early. This may be a drone for the escort, but the escort itself would be a picket for an aircraft carrier.

The best approach against sea skimmers is most likely to have airborne radars searching for and tracking these threats. The classic AEGIS system still couldn't shoot at them until very late (which means gaps in the defence if the escorts are widely spaced from the protected ships), but area air defence missiles with an illumination-independent seeker (or illumination by an aircraft) can engage targets beyond and below the horizon. The U.S. Navy very belatedly introduced this capability with the expensive SM-6 missile and will likely get ESSM Block II missiles (both have active radar seekers derived from the AMRAAM's) in a few years. The Italians, French and British use the Aster missiles, which have active radar seekers and can engage beyond and below the horizon if there's an cooperative engagement capability. Some offboard sensor (typically AEW radar) needs to provide the target data by datalink to the warship. The Israelis introduce Barak-8, also with an active radar seeker.

E-2D Hawkeye AEW aircraft. Capable, but also slow.
Such high end area air defence missiles are promising against the most challenging aerial threats, but their high price is a problem. An anti-ship missile may be cheaper than the missile launched to intercept it. There's thus a good case for a hi/lo mix with rather long range active radar seeker missiles at the high end and many smaller and much cheaper shorter-ranged missiles with a cheap guidance (could be radio command control as in RBS-23 BAMSE) at the low end. Alternatively, one could leave low cost threats to close-in weapon systems. Those should be mounted on all important ships anyway.

Even very expensive missiles shouldn't be expected to hit every time even under ideal conditions. The probability to kill was anything from about 0.9% to 99% with different missile types in military history. Sometimes missiles simply malfunction, sometimes the target escapes with evasive manoeuvres, sometimes electronic countermeasures interfere with the guidance and sometimes the fuse gets duped and the missile explodes early.
Nobody knows for certain how well surface-to-air missiles perform against an adversary anti-ship missile design, even if the latter was exported. This may lead to a repeat of the often-seen practice of multiple missiles launched on a single threat aircraft or threat missile. This tactic makes a lot of sense when an aircraft is in a box of SAM batteries that can shoot at it from different directions. A warship that's trying to intercept an incoming missile would use multiple missiles for intercept in order to improve the odds. A 90% hit probability SAM may provide an up to 99% defence when two instead of one were launched.
There's also the possibility of a sequential engagement, such as detection of a sea skimmer at 10 nm, first missile fired ASAP, second missile fired (if the first one missed) when the target is at about 5 nm, and the 3rd layer of hard kill defences would then be the CIWS (possibly missile-based as well).

To summarize this section about area air defences: Area air defences are first and foremost about having airborne sensors, for only then are air defences might stand a good chance to defend a convoy against sea skimmers.

Now keep in mind the intro to the series; warships should be used for missions on the high seas, distant from friendly forces on land. This means land-based AEW availability would either be sporadic or unaffordable. Naval AEW like E-2 Hawkeye (or the Chinese equivalent) could be available to a carrier battle group, but would likely be unavailable to mere cargo ship convoys. In the end, you better have AEW helicopters available. The Russians have one type, and the British have the most modern one (AW101 / CROWSNEST).

AW101 with CROWSNEST radar kit
Missile threats

I'll show a selection of representative or interesting modern anti-ship missiles (AShM) to clarify what kind of threat missiles warships may face.

Exocet - the classic sea skimmer
This kind of missile appeared in the 70's and became famous during the Falklands War. The missile is subsonic and it flies (for some time of the flight) at an extremely low altitude (sea skimmer); few metres above the water. The seeker is an active radar. This is the typical concept of Western anti-ship missiles. Ship-, air- and submarine-launched versions are available and launch from land is possible as well. The Russians later adopted this Western concept with the Kh-35, but missiles of this pattern are no more near the state of the art. Hardly any such missiles are adapted to launch from vertical launch silos, which leads to a widely known quantity of missiles carried onboard warship types (the launchers are unique to the missile type and visible).

NSM - the stealthiest
The NSM is a subsonic sea skimmer that doesn't use an active radar, but an imaging infrared (IIR) seeker. This makes radar warning receivers and most soft kill countermeasures ineffective, but the IR seeker can be defeated with multispectral smoke and dazzling lasers (DIRCM). The decontamination systems on warships (spraying the own ship with seawater to wash NBC contamination off) may help as well, as infrared sensors are not good at looking through water spray and the washing water would harmonise the ship's surface temperature with the sea. The same physics also affect the range of IIR sensors in rain, fog and when spray from waves is in the line of sight (which may force the missile to fly a few metres higher than otherwise desirable).
The pattern recognition capability of an imaging infrared sensor hints at the multi-role (including land attack) potential that the JSM derivative shall have. The missile may be able to aim at specific parts of a ship, such as the vertical launch silo. It's also thinkable that it can identify ship classes and select the highest value target. The NSM also offers some smart autopilot that can navigate over and around islands; that's almost a necessity in Norwegian waters.
The autopilot of a NSM may furthermore enable it to attack at dusk or dawn from the (approximate) direction of the sun, rendering the missile invisible to infrared and ultraviolet horizon scanners in addition to rendering ESM/RWR useless by not emitting treacherous radio waves towards the ship. It's thus very likely that NSM would be the missile type that is as of today the most difficult to detect, and would be detected at the shortest distance if it was employed well. The much heavier and more expensive AGM-158C LRASM may soon supersede it as the stealthiest anti-ship missile in service.

Hsiung Feng II - sensor fusion 
A subsonic sea skimmer missile with added infrared sensor that makes it much harder to deceive the guidance. This approach is not as stealthy as NSM, but it's even harder to defend against with soft kill countermeasures. The latest Standard Missile 2 area air defence missile as such sensor fusion with a secondary IR sensor as well. This approach is slowly becoming more popular.

Brahmos - barely "small" enough for a large strike fighter
The red missiles are Kh-31; "normal" size AShMs

Oniks/Yakhont/BrahMos - the Soviet big missile approach finally learned sea skimming
This is a large missile (about three tons weight) with a slightly heavier than normal-sized warhead (about 250 kg) and hundreds of km range. A top speed faster than Mach 2 is combined with a radar seeker and seaskimming capability.

A normal-sized missile that separates a second stage for a supersonic sea skimming terminal approach. The subsonic cruise is likely meant to improve the range over an equal-sized all-supersonic missile. Ship-, air- and submarine-launched versions are available.

two stage Club AShM
Hsiung Feng III - supersonic sea skimmer at normal size
This missile is a sea skimmer with supersonic speed, powered by ramjets. The French ANS project and the Russian Kh-31 are similar. Such missiles are similar to subsonic sea skimmers, but the about tripled speed leaves inversely proportionally little time for defences after detection.

BrahMos-II - hypersonic
An anti-ship missile (development project). This missile is meant to be hypersonic (supposedly Mach 5 and faster) while having a radar seeker and a useful range.

Marte ER - the small one
This missile as a whole weighs not much more than some other missile's warheads, but it can achieve ranges that put the platform out of the range of most area air defences. Such missiles can be launched from helicopters and just about any platform but submarines.

AGM-158 JASSM - the land attack cruise missile
Several types of air-launched land attack missiles (JASSM, Taurus, Apache, Storm Shadow) were created in the West during a period of hardly any (and even cancellation of) Western anti-ship missile projects. It's thinkable that such cruise missiles possess an anti-ship capability, albeit most likely without a true sea skimming capability. These missiles can at the very least be considered a threat while a warship is in port. They might also be expected to protect coastal regions against such threats.

IDAS - a tiny fibreoptic-guided submarine missile
This missile is mostly meant to defend submarines against ASW helicopters and small surface boats/drones, but it can also be used to damage surface warships where it hurts. A hit on a helicopter on the helipad or on a hangar would be especially troublesome despite the small warhead. The missile is very slow, but also has a small signature. Multispectral smoke would help little on its own, as they don't last nearly as long as such a missile could delay its terminal approach. Dazzling the missile's imaging infrared sensor on the other hand would be promising (and such DIRCM might be the only effective defence against it for ASW helicopters).

Kh-58UShK - tenacious anti-radar missile with a backup sensor
This anti-radar missile is supersonic, has a much bigger warhead (149 kg) than the Western HARM (66 kg). The new UShK version is configured to be able to home on a target even if it ceased to radiate, and applied to the naval context the secondary IIR sensor could be used to dismiss decoys of any kind and even seek to hit specific parts of a warship (such as main combat sensors, helicopter hangar, gun turrets or vertical launch silo). The Western AGM-88E combines a passive radar and an active mmW Radar instead.
Guided bombs and heavy unguided rockets
Aircraft could attack ships with immunity if there was no hard kill defence with a good effective ceiling.

Land artillery
Some land forces artillery can be a threat to ships, but most of these threats cannot reach farther from the coast than 100 km. The planned modification of ATACM and its successor missile is an exception to this rule.

IAI Harop - the kamikaze drone
This is one of many kamikaze drones - essentially guided missiles that are slow and have some endurance for loitering. It's small, accurate, has a passive radar and an electro-optical sensor and is remotely controlled. It ceases to be of utility once the radio link is jammed, but the man in the loop and the radio link can be replaced by computing power and algorithms, including pattern recognition for identifying ship classes and aiming to crash into the most critical vulnerable parts of a ship. These drones/missiles are a challenge because they are rather cheap and small and can thus be incoming in quantities much greater than the air defence missile supply of any warship. On the other hand, such drones/missiles are rather easy targets for guns and their very cheap munitions.

MALD / MALD-J - the free-flying decoy
Free-flying decoys have already been used during the Vietnam War to deceive and provoke North Vietnamese area air defences. The introduction of MALD long after the end of the Cold War signalled the availability of free-flying decoys that were small and effective enough to justify being used by strike fighters. Such decoys are no threat to a warship itself, but they may provoke it to deplete its limited inventory of area air defence missiles and possibly to activate its radars (revealing itself as the high value target among decoys and transport ships).
MALD-J is the decoy version that includes a jammer. It's thus more strike fighter-like and may interfere with the ship's ability to detect approaching missiles early.

S-300FM / SA-N-20 / Rif-M - long range air defence missile
This Russian long range air defence missile  may hit a warship at 150 km with a 150 kg warhead and would approach at supersonic speed.

ESSM - very short to medium range air defence missile
This rather small missile uses semi active radar (terminal) homing and proved its capability to hit surface targets as small as rigid hull inflatables at speed. The missile is quite expensive for its size and delivers a small warhead, so it's no substitute for real ASuW munitions. Its predecessor Sea Sparrow showed its dangerousness to warships in an accident. ESSM has no booster stage and so achieves a minimum engagement range of few hundred metres.It's a Western system and no threat to Westerners, but the concepts of both Block I and Block II versions can be expected to be in use in many non-Western navies.

Ballistic missile threats will be the topic of another blog post.

Terminal effects


(Sorry for the stupid rock music and the misleading still photo.
The video is a handy compilation of missiles impacting in ships in tests and exercises.)

The non-nuclear warheads of anti-ship missiles are much less lethal to warships than heavyweight torpedoes even with a similar explosive load. It's not really necessary to sink a warship, though. Often times that's an unnecessary cruelty, as severe damage suffices to put a ship out of service for the duration of a war.
A missile or air attack rather has to accomplish a mission kill. This may be a firepower kill (the ship being unable to defeat threats) or a mobility kill (the ship being unable to steer or to move at a sufficient speed). Even small missiles with malfunctioning fuses may cause a firepower kill by cutting a few cables in a radar mast.

Attack scenarios

There are three basic ways how to do a promising missile attack; surprise, saturation (brute force) and combined arms (strike package).

The aerial Exocet attack that crippled and sank HMS Sheffield in 1982 during the Falklands War was a surprise attack of a lone strike fighter. The unintentional hit on the USS Stark in the Persian Gulf was another (accidental) surprise attack; this target didn't switch on its defences in time. It would be possible to launch surprise attacks from Q ships - ships that were not identified as armed and hostile in time. This is particularly troublesome in maritime traffic bottlenecks if a convoy has to pass close to dozens of seemingly neutral ships that it cannot possibly board and inspect in time. Anti-ship missiles may then be launched from a very short distance, similar to the INS Hanit and HMS Glamorgan in the Falklands War. These cases also support the notion that land-based missile batteries are important for coastal waters warfare - and this extends by hundreds of nautical miles offshore with long ranges of modern anti-ship missiles.

The Soviets planned for saturation attacks with dozens of missiles incoming at one time, which was promising back when CIWS were usually ineffective and area air defence missiles had to be guided by the ship's radars illuminating the target with a radar beam. Only one target per illuminator radar could be engaged at a time, which meant that a typical frigate could hope to intercept a mere one or two missiles before they arrived.

Combined arms or strike package attacks are the most demanding ones, albeit you can in theory pack the ingredients into a single missile (if it deploys dissimilar submunition missiles). Some aircraft may chase off or kill an AEW aircraft, another one may jam radars, a third may jam communication datalinks, yet another may sent anti-radar missiles to a ship and a fifth may time the launch of its anti-ship missiles so they arrive while the incoming ARMs almost compel the target ship to shut down its radars. This can be enlarged to strike packages of dozens of aircraft.

Soft Kill

"Soft kill" includes the means that avoid a hit without destroying the missile. Deception, disruption, concealment and (borderline) dazzling fit into this category.

Some soft kill measures can be deployed all the time (such as a towed radar decoy), while others become effective only upon activation. Most soft kill countermeasures do thus require a timely detection of an incoming missile threat. It may help to determine the type of incoming missile, and to understand its guidance, for else you'd have to expend munitions for countermeasures just in case, even though  some of them don't affect the missile. Even worse; one countermeasure to a simple anti-radar missile is to switch off the threatened radars. An imaging infrared-only missile defence wouldn't be affected by this, but radar-dependent hard kill defences would become unavailable.

A large quantity of soft kill measures against missiles exploit the stupidity of the missile's onboard logic (computer), which is increasingly difficult. This is much harder to do with a man in the loop (remote controlled missiles) - in that case you better disrupt the guidance method, such as deploying smoke as concealment against semi-active laser guided bombs or jamming of a radio link.

Pandarra fog experiment, USN 2014
Radars appear to be particularly prone to false targets because they are usually very poor at creating imagery of objects that could be used for pattern recognition. Shorter wavelength radars such as the very short-ranged millimetre wavelength radars can create such imagery and the addition of a high resolution mmW radar to a longer-ranged search sensor has likely a similar effect as the addition of an imaging infrared sensor.
To defeat such a sensor fusion or two or more sensors is particularly difficult because deceiving both sensors might still not suffice, as the only commonality between their returns (real and false targets) may be the real targets.

There are several types of radar jamming and I'm not nerdy enough to go into greater detail about them than even wikipedia already did. There's a series of books called "EW101", "EW102" etc. that hold enough details for more inquisitive minds.

AN/SLQ-49 inflatable corner reflector
Old (1960's and 1970's) radars were and are easily deceived and lured onto false contacts with things as simple as corner reflectors (metal surfaces in 90° angle produce an extremely strong radar return), chaff clouds and even helicopters were able to pose as a ship by hovering low and to escape from a missile by climbing (the missile was programmed to fly low when close to the contact).

Radars and attack logics have become more sophisticated (though a 10 year old missile may lag behind desktop computer tech by two decades), and so have radar decoys. One rather sophisticated radar decoy is a free-flying rocket-driven hovering active radar decoy. Offboard radar emitters emitters have the advantage that they may lure missiles that decide to follow their passive radar's data.

(MK-53 NULKA radar decoy in action)

Radio link jamming may be done with radar jammers if the radio wavelength is covered. There are hardly any missiles in use with radio command guidance (particularly not as anti-ship or air/ground munitions), but radio jamming may become more important when incoming missiles cooperate with datalinks to fuse their sensor data from different angles into a common picture of the situation. Missiles that fail to get reliable distance readings might triangulate distances if they cooperate, for example. A disruption of radio datalinks may be of great utility before the missile has a line of sight to the real target. This kind of radio jamming could be done by helicopters or drones. A missile may thus be blocked from receiving targeting updates from manned platforms (such as a launch aircraft). For example, such targeting info updates might inform it about two ships having switched their position in a formation - and the missile might thus go after an unimportant decoy container ship instead of an aircraft carrier.

infrared imagery


Remember; a computer can make use of the raw data, while humans look at images
created with raw data and different settings. Algorithms know
about tiny temperature differences while we don't.
Infrared countermeasures cannot really create effective false contacts against imaging infrared seekers. One might have a miniature of a ship with identical shape and temperatures and try to fool a missile into thinking it's a more distant real ship, but moving to right and left enables enough triangulation to rule this out unless the decoy is rather big. The growth of the silhouette size as the missile approaches at a known speed also allows for calculations that reveal the ruse. Old radars saw both ships and chaff clouds as similar points, but imaging sensors can use pattern recognition, and a multispectral chaff cloud doesn't look like a ship at all. Such smoke clouds can still conceal a real ship, and challenge the missile to guess where's a target behind that wide cloud. Such a multispectral cloud cannot be maintained for long, of course. A ship can easily carry and launch munitions for a few minutes if the crew is prepared for  it, but neither chaff nor multispectral smoke can be maintained for hours, and the more wind there is the shorter the duration of the concealment.



Water spray and cooling the ship with seawater to the water surface's temperature might help against IIR seekers.

A dazzling laser may defeat an IIR seeker if it's in the seeker optic's field of view, and with enough power it could even damage the optics permanently. Laser dazzlers may be ineffective against a missile that avoids looking at a laser-equipped escort and instead attempts to slip past it to the escorted transport ships of a convoy.

In the end, there's a huge problem with soft kill countermeasures: They provoke a miss by the missile, but they don't get rid of the missile unless it crashes into a decoy ship or attempts to dive onto the target. Some missiles are 'smart' enough to simply turn around and try again, and again and again.

Close-in weapon systems

The most famous close-in weapon system is the Vulcan Phalanx, but it doesn't deserve this fame, as it's also one of the lousiest CIWS (poor effective range). I will instead present four CIWS that represent different and modern approaches to the problem of destroying an incoming missile early enough to avoid its wreckage badly damaging the ship anyway.



Heavy (initially 75...88 mm) anti-air guns were the earliest dedicated anti-air weapons. The earliest examples were even called "Ballonabwehrkanone" in Germany - as they were meant to shoot at balloons, not heavier-than-air aircraft. Heavy AAA (anti-air artillery) was still very important in the Second World War, but they were truly efficient only with smart fusing solutions (dual time/impact fuses or proximity fuses) or when the targets flew in a stupid way (straight line, only slightly higher than 4,500 m). Fire control was most difficult before radars and computers became available.
One of the conclusions of WW2 was that 76 mm AAA was the way to go, for it allowed for many rounds per minute with good effect using proximity fuses. This conclusion was only valid up to a disappointing range of few kilometres, though. The high speed of turbojet aircraft made their evasive manoeuvres too effective at long ranges.
76 mm guns still made a comeback during the Cold War as naval guns with emphasis on air defence, in between smaller calibre CIWS and missiles. They were typically used instead of rather than alongside heavier guns (the Italians were an exception). 76 mm guns became popular as main and sole naval gun on frigates during the 70's. They were much more lightweight than 100 or 127 mm guns and provided some versatile gun firepower as backup to other wise totally missile-dependent warships. The Italians supplied these 76 mm guns and developed them up to 120 rpm rate of fire (76 mm Super Rapid) and gave them reduced radar cross section turret shapes.
The accuracy and thus effective range can be improved with radio beamrider (guided) subcalibre shells (DART; to approx. 6 km).

The RAM began as a quite cheap infrared-homing missile with a better range than the then-typical Gatling-based CIWS. It was upgraded and received an additional passive radar homing sensor. The hit probability in peacetime tests is near-perfect, but this missile is 100% useless against a NSM approaching from the direction of the sun. Moreover, the missile has become expensive, approaching the price region of area air defence missiles. Personally, I don't think the RAM can be trusted as sole CIWS despite its stellar reputation.



35 mm Millennium Gun
This 35 mm revolver gun achieves a rate of fire of 1,000 rpm and combines the external ballistics of a high density 35 mm shell (much better than what can be had with full calibre 20-30 mm HE munition) with a cannister/shrapnel round combination marketed as AHEAD. The spinning 35 mm shell opens up after a pre-programmed time and releases many tungsten pellets. They disperse in a near-circular fashion driven by the centrifugal force. These pellets form a sufficiently dense cloud only if the timing of the release was correct. The use of such a CIWS thus requires correct range information not only for aiming the gun, but also for programming the fuses. The smaller the target silhouette, the more challenged the system is - evasive manoeuvres by the missile on the other hand don't look very promising below 1,000 m distance to the ship (they would even enlarge the target silhouette).

pellet pattern after a burst of several AHEAD shells
The advantage of such a system is that it would be most cost-efficient when used against swarms of cheap flying drones.

The Barak-1 missile is a bit more capable than the older Sea Wolf VLS missile. The missile can be produced cheaply (it's not necessarily sold at a low price, though) due to a simple guidance that depends on the ship's radar (no seeker in the missile). The effective range is short, but better than with all guns.
 
Problems with CIWS

The primary problem of CIWS is that few seconds are available for the intercept, and much fewer seconds against supersonic missiles. There would likely be no point in trying to use CIWS against a hypersonic missile.
CIWS have a hard time shooting down an incoming missile that's been detected but a few seconds prior to impact. They have it much easier if said missile was lured to approach a decoy at first, and passes by the ship. The missile may then turn for a new terminal approach, but this still buys many valuable seconds for the CIWS and the target silhouette of a turning missile or drone is much bigger than the frontal silhouette. A CIWS may thus make sense even if it stands no chance to intercept a missile on its first approach.
Another problem with CIWS is that they're often simply shut down when a missile approaches. So far CIWS have scored no missile kills at war - but Phalanx shot down a friendly aircraft in an exercise and another time shot at chaff when a real missile was incoming. Meanwhile, soft kill defences have defeated about half of all anti-ship missiles launched post-WW2.

Yet another problem with CIWS is that the ship may be damaged even if they score a kill. The wreckage of the incoming missile is a threat for several hundred metres after the kill.

CIWS typically require emitting sensors (radars), most CIWS are utterly unable to function without this. Such radar emissions can be picked up by a missile and used by its target engagement logic, even to dismiss non-emitting chaff in favour of attacking the real ship.

The Argentinian bomb run at San Carlos bay in 1982 that the area air defence technology of the time could not defend against pressed home the insight that CIWS are a necessity. The Argentinian pilots could not be fooled by chaff (and smoke wasn't deployed in time). Maybe this reasoning should have changed when non-line of sight area air defence missiles such as Aster became available. Such missiles can defend against terrain-following aircraft and missiles over land if they are fed the necessary targeting data (possibly based on aerial radar data).

Some more remarks

Warhead and fuse concept depend on the target type. Some missiles are built for hit-to-kill and carry no or only a tiny warhead. Such missiles are obviously ill-suited if not entirely useless against very small and possibly at the same time very fast targets such as anti-radar missiles. Expanding continuous rod warheads are fine for cutting a strike fighter in two pieces, but getting the timing right to hit a short and fast missile with it would be too difficult. A fuse that's got the right timing to hit a target with an approach speed of Mach 2.75 (Mach 2 interceptor + Mach 0.75 subsonic missile threat) may have a poor timing for hitting a target at an approach speed of Mach 8 (Mach 3 interceptor + Mach 5 hypersonic missile threat). The optimal fragmentation patter of fragmentation warheads would be different as well. Long story short - there are versatile missile designs and missile designs tailored to specific target categories.

IIR seekers may have an easier time to lock on missiles with low radar returns than radar seekers, but they're unsuitable for Mach 3+ missiles over long ranges because the window heats up and blinds the seeker unless there's a removable protective cap (example IRIS-T SL).

There is a hypothetical missile threat of a subsonic missile that deploys free-flying decoys or even worse - faster anti-radar missiles (ARM). This would make it easy to achieve the correct synchronisation of anti-radar missile attack and anti-ship missile attack and would for the warship to actively defend against the ARMs instead of sitting them out silently.

The classic problem for area air defences is that they often have huge missile expenditures and run out of missiles (it happened to the North Vietnamese repeatedly, especially with the extremely low accuracy SA-2). This problem was on the mind of the British when they stretched the Sheffield destroyer class into the Manchester subclass in part to add 15 more Sea Dart missiles. Quad-packing (ESSM) and stacking (Barak) of missiles in vertical launch silos helps to get the most out of limited VLS capacity, but there's still a good reason to believe that VLS sizes are too small on many frigates: Missiles are damn expensive, and it's every navy's instinct to want nice and many ships more than hidden munitions. Most navies appear to have little if any reserve munitions - they aren't necessarily able to even only fill the magazines of all their warships.

This graphic shows a diagram about effective ranges of a short range surface-to air missile. The missile isn't terribly fast and a beam rider and thus the effective range depends greatly on the velocity and pass distance of the target. It's a bit extreme in this regard. The lesson one can draw from this is that one shouldn't take the claimed SAM range and then make conclusions about how many escorts would be needed for gapless area air defence if they formed a ring around a convoy. That's not how it works - the effective ranges are much worse, even disregarding the detection and tracking challenges.

Some air defence modules may be bolt-on modules, while others are through-deck modules. The former are more easily placed, especially as upgrades. The latter have more potential for ready munition.

Over-the-horizon radars are in existence and have been in existence since WW2, but they don't really help with the horizon problem. They need large installations to work well and to disperse the antennas to multiple ships is impractical. Moreover, they are blind on the first few hundred kilometres distance - even if they're able to detect an aircraft a couple thousand kilometres away. The illustration here shows this effect.

Cooperative engagement seems like a self-evident thing to have - digital datalinks have existed since the 1960's already, after all. It's apparently still not commonplace, and that's the nice choice of words. The USN itself is lagged behind its AEGIS propaganda of the early 1980's by almost a quarter century; that's how long it took to hit a target with a missile without the launching ship having a line of sight to the target. AEGIS propaganda created the impression that such coordination was possible during the Cold War already.
We cannot assume that beyond and below the horizon target can be engaged even if a warship has missiles that could engage it and an AEW is on station. The fire control chain may still be broken.

Redundancy is a promising approach to make a convoy's defence more robust (fashionably: "resilient"). U.S. Navy guided missile cruiser designs were double-ended; the same weapon systems were on the forecastle as on the afterdeck. There were multiple illuminator radars, but no redundancy in search radars. Another way to achieve redundancy is to simply have more warships than bare minimum. Either way - redundancy is an expensive insurance.

SPG-62 = radar illuminators for SM-2 and ESSM Block I

Electronically scanned radar antennas have many advantages, and have been popular at sea since the 80's, but they are still limited by radar physics. The greatest advantage provided by these radar antennas is likely that a horizon sweep happens at much shorter intervals. This does hopefully lead to a slightly earlier detection of an incoming sea skimmer. Another advantage of AESA radars in particular is that they can compensate for failing antenna modules. Damage by fragmentation is unlikely to put the entire radar out of operation, unless it cuts the cables to the antenna. AESA radars also tend to be more multi-function than more old school rotating antenna radars, which makes a difference in particular on small and not lavishly budgeted ships that can now make do with a single radar for navigation, surface fire control, horizon scans, volume search, air target fire control and air traffic (helicopter) handling. Even those should still have a civilian navigation radar as backup, though.

Warships are not in 100% condition all the time. Electronic components have mean times between failures - certain radar components might fail after a couple thousand hours of operation. That's one reason why they aren't in operation all the time. Hundreds of critical parts could fail anytime and make a warship much less capable in AAW. A transoceanic convoy action or two in quick succession may  lead to many subsystems failing on a ship even if it was in 100% condition on the first day of action. Some failures would be repaired (often by replacing parts) in minutes or hours, others would only be repaired in port. Some major failures may even require repair in a shipyard.
Warships are thus not going to be fully capable all the time, even if the crew was perfect. This might punch temporary holes into any air defence of a convoy, which is yet another good reason for redundancy.

Radar stealth is helpful at long distances, but any missile that had good-enough targeting data to reach the vicinity of a real target will still find that target regardless of its radar stealth. In theory it's even possible that a missile could climb after being detected and look at the sea surface to spot ship's the wake pattern to tell real ship targets from decoys and boats or to sense that the echo of the sea surface is different or missing in one place.

High-powered lasers may in theory heat up a warhead till it explodes or melt a radar antenna. Such lasers are being discussed and pursued for shipboard use, but they are rather unlikely to become important soon. They need a line of sight, which leaves them but a few seconds to intercept supersonic sea skimmers that are already made heat-resistant to withstand the heat generated by friction as they move at supersonic or even hypersonic speed. Most military explosives are "insensitive" today, which means that you need to heat them up to much higher temperatures than earlier explosives before they react. I suspect lasers will be used for dazzling and damaging IR optics, while more powerful lasers are likely not worth the effort for a long time to come.

Radar physics are more complicated than I indicated so far. My previous repeated mentioning of horizons was a simplification.
Sometimes a sea skimmer flying at 10 m would not be detected by radar while one flying at 4 m altitude would be detected by the very same radar. Radar waves don't have the geometrical or optical horizon; the radio horizon applies, which is different depending on the wavelength and atmospheric conditions. Some radar wavelengths cannot be used to detect high-flying aircraft, some wavelengths have ground propagation or are reflected by upper layers of the atmosphere. Shipborne radars do have a radio horizon, though - it depends on antenna height (mast size), wavelength and atmospheric conditions.
 
Radars have a nominal range, but do not have a 100% probability of detecting a target at that range. Jamming, atmospheric phenomena and radar cross sections (RCS) at the radar's used wavelength have influence and may lead to much lesser detection ranges. Almost all not very low RCS objects in line of sight should have been detected at about 1/2 to 2/3 of the nominal range as a rule of thumb. Additionally,  the whole (radio) horizon issue makes nominal ranges quite moot.

Aircraft can deploy towed decoys that may not only lure a missile that's approaching from a side, but may also trigger the missile's fuse and thus its self-destruction at a safe distance from the aircraft. So far I have not seen any indication that such tech is in use to enhance anti-ship missiles' survivability, but it's not unthinkable especially for subsonic and larger missiles. The missile would need to deviate much from a straight approach to the target to bring a towed decoy into play, though.

Smoke used to be an effective screening method against naval gunfire, but nowadays it's questionable if effective at all. It's blinding the defender just as it blinds the attacker (or its precision guided munition). The problem is that most anti-ship missiles would simply fly through the smoke without fusing and thus notice that they chose the wrong target - and then they may choose anew, maybe a real target. This buys time for hard kill defences, but the smoke may also keep the same from scoring a missile kill. Ideally one would identify an incoming missile type and deploy a tailored smoke as concealment. So missiles that rely on radar would be countered with infrared-transparent smoke which allows infrared-based fire control for CIWS. IR-dependent missiles would be engaged with IR smoke that still allows for radar-based CIWS operation. Against sensor fusion missiles - well, here we see a small part of the appeal of those.

Conclusions

First and foremost, the air defence of a convoy should have AEW support, be it with a fixed wing or rotary wing platform, manned or unmanned. There's no more than short range defence possible against most modern threats without AEW support. And by "AEW" I mean airborne radars with good enough sensor to provide fire control quality data, not just an indication of the approximate location of a threat.

The ranges of AShM types have become so great that an attacking aircraft can launch the missile at a safe distance, and a AShM salvo may even be launched by so far undetected submarines. The focus on intercepting the munition rather than the platform is thus correct.

Even a handful of SM-6 can cause major restrictions to opposing air power (especially with the quantity in VLS being kept secret), but their high price and large size (one missile per VLS cell) makes them unsuitable for the role of primary area air defence missile.

ESSM provides great packing density (four missiles per VLS cell) and the Block II version with active radar seeker is on course to become a standard area air defence missile. The range is much better than with CAMM and VL MICA, and the superior packing density and much shorter minimum engagement range (both due to using no booster) outperform Aster. ESSM Blk II will require no dedicated illumination radar and in fact not even a powerful fire shipboard fire control radar. It has a semi-active mode that could help with choosing one target in a tight group of targets, but its use is optional.

Shipboard sensors should excel at detecting sea skimmer missiles at the horizon (with radar and IR sensors) and at even shorter ranges, but the great expenses that powerful area search and long range radars require are dubious in light of the utter necessity of AEW. This adds a huge bold question mark behind the concept of dedicated AAW frigates and destroyers, as said powerful radars are their primary characteristic.

ASW capability is a necessity on every dedicated warship, short range AAW is a necessity as well, and participation in AEW-supported area air defence AAW soon requires little more than the addition of a few vertical launch silo cells loaded with ESSM Blk II. This leads to my conclusion that future dedicated warships should be general purpose warships (GP, ASW + AAW).
There's no way how having separate AAW and ASW warships could be the more economical solution (same naval capability for smaller budget) when all it takes to turn ASW unit into a GP unit is to add a few VLS cells with active radar seeker area air defence missiles (preferably ESSM Blk II quad pack and SM-6).

The close in defences need to move towards a greater emphasis on the ability to kill many low value drones / missiles. Approaches such as the RAM missile are unaffordable in this regard. The 35 mm Millennium gun is a promising weapon for CIWS if the ship has a main gun of 100 mm or bigger calibre, and the 76 mm Super Rapid with mostly optionally proximity-fused HE shells and a few DART rounds is a promising armament for a ship that's got no bigger gun than 76 mm anyway. Both use relatively cheap munitions and could intercept cheap AShM / drone types at useful distances, even though their ability to protect against by high end missile attacks is very questionable and may be negligible (especially regarding hypersonic missiles).

Soft kill methods are relatively cheap and some of them are extremely cost-efficient. Soft kill defences should be on all ships in threatened regions - both military and civilian ships. Dazzling lasers for defence against imaging infrared-guided missiles should and I suppose will become commonplace at least on warships.

The defence of a convoy needs some redundancy built in at several levels in order to be worthwhile in face of capable adversaries.

- - - - -

I still didn't cover everything on modern and near-future AAW at sea. Hardware was presented in representative examples only, radar physics were barely mentioned, aspects of crew training/quality and readiness in general were barely touched. This part should still suffice to draw conclusions on what force development paths Western navies should choose (part VII).

S O
.