I recently read a guest post by Jonathan La Force on the difficulties the PRC would face if it invaded Taiwan. This blog post is an extension of the comment I made to that excellent post.

Map of Taiwan and vicinity (from google)

To invade, the PRC has to get an army across the ~200km (100 nautical miles /125 miles) separating Taiwan from the mainland (see map above) and defeat the government and military once it has done so. Mr La Force concentrates on the journey, which is challenging – and seasonally limited (see below) – I’m going to talk more about the destination. I’m also assuming that, at least initially, Taiwan is on its own and the US and Japan do not get overtly involved, his article explains the wet dream that a Taiwanese invasion fleet would be to US Navy surface and submarine assets.

To start with the obvious. Taiwan is a lot smaller than the mainland. It is some 80-90 miles wide at the widest point and 230 miles long. Taiwan’s GDP on a PPP basis is a twentieth of the mainland’s and it’s military budget (US$17 B according to wikipedia) is less than a tenth that of the PRC – US$229.4 B (again according to wikipedia). It has a population of some 23 million and the male population in the key 20-50 age range is under 5 million. This means that there are far more PLA active and reserve force members (2.5M active 5M reserve) than there are Taiwanese who could be sensibly recruited to fight. The Taiwanese military has an active force of some 165,000 and a reserve force of about ten times that so trained personnel numbers are even more in favor of the mainland. The same goes for ships, aircraft and so on. In other words at the top level, the invasion looks like a slam dunk because the difference in sizes of the two nations economically and militarily is so overwhelmingly in favor of the mainland.

But let’s not forget that not all islands are the same and it gets better (or worse if you are the PRC) when you get into the details of geography and meteorology.

Taking the latter first, there are certain times of the year when you don’t want to invade lest your fleet imitate that of the Mongols almost 800 years ago when they tried to invade Japan. Typhoons are not as dangerous to modern shipping as they were to the wooden ships of the 13th century, but they can still ruin the loading and unloading of military equipment and seasick troops are unlikely to be at their prime either. Plus even if you land before a hypothetical typhoon, if one shows up while you are invading it’s going to put a major crimp in your logistics. So you’ll want to avoid invading during typhoon season – late June to early October – and probably it makes sense to skip the immediately prior “rainy season” because large chunks of military logistics are far better done in the dry. Since you probably also want to avoid winter storms that means you only have late October, November and mid-March to mid May as suitable times to invade. The preferred time of the year is almost certainly late October to November because you can prepare around the typhoons and launch as soon as you are sure you have no typhoons for the rest of the year. Now this sort of limitation doesn’t necessarily impact the overall success of the invasion, but it does mean that the defenders have periods when they can safely take equipment off line for maintenance and replacement and so on.

That leads to a related point, there is no way that in these days of ubiquitous satellite observation a maritime invasion is going to be a surprise. The build up of ships in the ports of Fujian is going to be completely unmistakable, but you can’t avoid doing it by splitting up the starting locations. A 200km crossing is 5 hours at 20 knots and 10 hours at 10 knots. That’s a long time for soldiers to be in the ships but trying to disguise the invasion by staging some of it from further away is to run the risk that your troops will be unable to go into action when they arrive (or even disembark successfully). The only way to compensate for this is air assault in the first waves but that too has issues. In order for the majority of your air assault troops to survive to start their assault you have to have achieved air superiority over the entire flight path and, while the required preparations to achieve that are not quite as obvious as gathering hundreds of ships, it is also highly likely to be spotted by satellites. The lack of surprise means that the defenders will be able to mobilize fully and to prepare surprises for your invasion force. That mobilization and those surprises are going to be assisted by the geography of the island which limits avenues of attack.

Geography

So let’s look at the geography. Taiwan is basically a big mountain range with a skirt of flatter land around it (more on the West side). It could hardly be designed better for resistance operations – something that every single invader in the past 500 years has discovered – the Dutch, the Imperial Chinese, the French, the Japanese, the KMT… Now eventually most of these invaders succeeded, but they didn’t when someone else was supporting the resistance, and even when the resistance was on its own it took a while and a fair number of corpses. So even if you can get a foothold on the island, expanding to get the whole island is going to be a long drawn out process.

But don’t worry, getting the foothold is also impacted by geography as even the skirt of “flatter land” is distinctly bumpy. One of the questions is where to invade? well, as we see from the map above the northern part of Taiwan is slightly closer to the mainland, and it includes the capital city Taipei, so almost certainly that’s the place you want to start with. If you don’t, you will at some point have to get to Taipei and your invasion force is going to face similar challenges as it would if it went for Taipei first, so this article is going to look at the direct Taipei invasion and it is left as an exercise for the reader to see how an invasion elsewhere would fare given the similar geography etc.

Taipei area map (from google)

Take a look at the surroundings of Taipei on this google map and note the crinkly green bits. You will not be driving or marching in from the north or south because there are 3000ft (1000m) plus mountains in those directions. Those mountains are steep and forested. Yes there are passes and there are rivers but you don’t get overwhelming force by driving in single file along a river valley or zigzagging down a mountain. East is marginally better – the hills are lower – but there’s still only one practicable route – the Keelung River and motorway to/from Keelung city. This assumes you land your invasion force at the Keelung port which seems kind of silly since you have to go around the top of the island to get there and that loses the element of surprise. Moreover, not only is it likely to be well defended, but Keelung port is a major container port and just scuttling one container ship in the harbor entrance is going to block the whole port. In fact you don’t need to scuttle it to impede access, just anchoring it across the entrance will be quite enough because the entrance is (according to Google maps) about 250 meters wide and large container ships, such as the notorious Ever Given, are up to 400m long.

An Evergreen Marine Corp cargo ship is anchored at the Port of Keelung –
https://www.taipeitimes.com/News/biz/archives/2021/05/10/2003757117

Mountains

In the photo above, as well as looking at the size of the vessel, pay attention to the hills in the background. These are typical of the area, indeed most of the island. They are steep, lushly forested, and will contain hiking paths, logging tracks etc. that defenders will know all about but attackers won’t. Tanks, mechanized artillery and the like are going to be of limited use to attack through them and they are almost ideal for defenders to setup ambushes, and deploy mortars, anti-tank weapons etc. to drop death on those on the roads below and so on. Taiwan does of course have highways (and railways) that cut through these hills with tunnels and viaducts so the defenders can move reinforcements and logistics fairly fast to the front lines, but those highways can also be easily cut by collapsing those bridges and tunnels. A 100m/300ft hill doesn’t sound like too much, and unless you’ve seen the forest in question it seems like you ought to be able to at least send infantry through the forest without too much trouble. But a lush semi-tropical forest is very different to the forests of higher latitudes and a 10-20% grade is a lot harder to climb than an 5% grade. The picture below spells it out even more clearly: if the tunnel (or viaduct or both) is destroyed you have to go through the steep bumpy forest instead and that’s simply not possible for vehicles until you have some combat engineers (re)build a road, and that requires that the surrounding area be clear of defenders.

A Taiwanese highway – https://goo.gl/maps/kBthDKV1FTRyaEE8A

The hills are the key difference between fighting in Taiwan and the last large scale East Asian war – Vietnam. If you go look at Google’s topo view of either Hanoi or Ho Chi Minh (Saigon) you see both are in flat parts of the country – there are some isolated hills around Ho Chi Minh city and there are mountain ranges relatively near Hanoi (~20 miles/ 30km) but in neither case will those mountains stop you driving tracked vehicles anywhere you feel like, presuming you can handle the rice paddy mud. You can tell that this is the case by the fact that the roads have lots of straight lines. Near Taipei away from the middle of the city straight roads are rare. If you look at Taipei and the surrounding New Taipei cities on the map, there’s only really one place where you can go East-West more than 10 miles in a straight line and not need to ascend a hill to do so. There is a big river in the middle of that so the furthest you can stay dry and go basically east-west is no more than 6 or 7 miles. North-south is even more limited. In fact there are lots of directions where traveling in a straight line from the middle of Taipei results in you going up or down slopes of 20% or steeper within 5km/3 miles, so even if you somehow manage to clear the forest (wild uncontrolled spraying of napalm perhaps?) you are still going to face challenges getting around. And, as I suggested earlier, Taipei is not particularly unique in this aspect, most of Taiwan is mountainous. Although the coastal plain south of Taichung all the way down to Kaohsiung is the flattest part, it’s only 30 miles wide at best so almost all of it is in artillery range from the hills to the east. In that part of the island, if you invade there,you have the rice paddy mud problem, and you still need to get to Taipei to convince the world that you’ve conquered the island. Only now you have even more mountains to get through to do so.

Rivers

Traditionally you can use river valleys to travel along roads that are reasonably flat. The problem with this regarding Taipei is that none of the rivers have the helpful flood plain bit where you have lots of flat land. All of them, in fact, have spots where they pass close to hills that can be used as bases for suppressing fire. Once you are sure you have cleared those hills of Taiwanese forces you can use them, until you have done that you are just asking for attacks by Taiwanese groups with handheld anti-tank missiles (the Javelins and NLAWs so effectively used in Ukraine) or spotters for more distant artillery to take out your attacking forces. In addition these rivers are surrounded by buildings and the roads have bridges over smaller streams. As we have seen in Ukraine taking down a bridge or two or collapsing a building onto a road makes challenging obstacles even for tracked tanks and they can make the route unusable for wheeled supply trucks. And you really don’t want to try sending boats up them either, they are just as vulnerable as the land vehicles to death from above plus there’s the danger of mines and other obstructions in the river. Plus, of course, if you drop one of the bridges going across the river you will make the river unnavigable until you manage to clear the bridge. The Tamsui river down to the sea can be blocked by dropping the Guandu bridge and clearing that, coincidentally, exposes your forces to death from above from hills on both sides of the river. In fact it is easy to see how the Taiwanese defenders would drop not just the river bridge but the clover leaf intersections at both ends and block the way to/from the coast for both land and water travel. Similar actions can probably block the upstream end of the river near the border of New Taipei and Taoyuan although that’s harder to do as the river is in less of a gorge. But the river is going to be very hard to cross if a few bridges are missing and both banks are overlooked at various points by hills that can conceal defenders.

What this means is that unless you achieve a successful air assault into the middle of Taipei really the only way to get an army into the city is via Taoyuan and most likely the banks of the river. This is going to be messy. So how about the air option?

Air Assault

The two airports near Taipei – which are the ones you have to use for an air assault assuming you want to conquer the capital city – have nice little mountains overlooking them.

The downtown one (Songshan) is so obviously going to be a deathtrap, you aren’t going to land troops there. Even if you land a couple of commercial airliners like A380s full of troops ready to go as a surprise attack, you won’t get any more assuming the Taiwanese have any kind of AA fire nearby. And even then, if there’s any artillery at all in those hills they can take out anything in or around the airport with one ranging shot, an easy visual confirmation of hit and then full battery fire. Seriously go look at google maps with the topo view, and feel free to zoom in to the bit near the google pin in the image above. You will note that there’s a hill to the north that is over 300ft/100m high and less than 2miles/3km from the airport. Plus the same hypothetical artillery in those hills can also shell anywhere in Taipei because pretty much the entire city is within 5 miles of them so even if you break out from the airport you can’t escape the artillery. And no you can’t land some paratroopers in those hills to take out the defenses. Well, you can land them, but most of them are going to end up dangling from the tops of trees with broken limbs so they won’t be doing a great deal of fighting because those hills are forested (see google maps satellite view and street view of some of the small roads). The British SAS tried parachuting into the jungle during the Malayan Emergency and it didn’t work for precisely those reasons, and the SAS didn’t have to face people shooting at them as they floated down. PLA paratroopers are likely to be shot at, presuming they make it out of the plane before it is shot down. You also can’t get some of your first wave of airborne assault to charge up the hillside – there’s a 400ft/120m wide river in the way with 1 (one) bridge. If the hypothetical artillery don’t already have that bridge locked down with precise ranges etc. (and/or the bridge has not been wired for demolition) they are demonstrating practically Russian levels of incompetence and that seems improbable. Also once your troops have got across the river they have to climb the hillside, which is (see above) forested and not notably filled with pathways and trails once you get past the built up lower parts.

So taking Songshan and trying a decapitation strike on the Taiwanese government from it is pretty much doomed to failure. It will make the Russian adventures in Hostomel look successful.

It is way more sensible try for the further one, Taoyuan, which also has hills a couple or three miles away but isn’t quite so obviously a death trap. It also makes a tad more sense, since it’s right by the coast so potentially it can function as part of the beachhead for the naval invasion force. As the missile flies, Taoyuan airport to the middle of Taipei is about 30km/19 miles so that seems doable in principle, even though you probably can’t get there using the same route as a missile but have to go south to the Tamsui river and then follow that downstream. The coast, as illustrated in google streetview, is flattish muddy beach with no obvious mine fields, tank traps, or other obstructions so a logical invasion would have an airborne force land at the airport, hold it for air support and have part of the force head to the coast taking out any resistance so that shortly afterwards the main invasion force can make a landing there.

In order to do this you need air superiority so you need to have taken out all the Taiwanese airfields, static radars and SAM batteries. A large missile barrage should suffice and the PRC almost certainly has enough working missiles for this – unlike Russia. Now if you’re lucky that’s going to be sufficient. If the Taiwanese have been paying attention to Ukraine (and if they haven’t then they deserve whatever they get) then they will have lots of mobile SAM batteries, radars etc. stashed in useful locations that won’t be taken out by the missile strike. Those will then pop up to take out your airborne assault unless you are able to confuse the issue with lots of other aircraft. This is, as with the missile strike, almost certainly doable. It will be expensive in pilots as well as in aircraft and missiles but it should be possible to keep a corridor over the strait of Taiwan and around Taoyuan generally open to aircraft if enough fighters, missiles and the like are thrown at the problem.

Taoyuan to Taipei

So let’s assume that you manage somehow to take the airport, and make a landing nearby and put several thousand troops on the ground. Lets also assume you have somehow got local air superiority and decent naval resupply. Even with all that, you now have to fight either through some steep, forested, hills to get to Taipei or go south through built up bits of Taoyuan, in range of artillery in those hills, all the way to the big river and then go down that to get to Taipei, and going down the river you are also, guess what? in range of artillery in the hills either side.

Now with some level of air superiority you can also bomb Taipei and perhaps the rest of Taiwan and hope that this causes the government to decide to surrender. But if the government doesn’t then the bombing is only going to make your advance into Taipei harder. You might want to consider the Russian experience in Mariupol and just how many troops they have had to use to subdue and mostly conquer that city. The greater Taipei metropolitan area (Taipei, New Taipei and Keelung) has a population of 7 million compared to Mariupol’s prewar population of about 200,000. And Taoyuan itself adds another couple of million to that count. If many of these people want to fight there will be an unending guerilla war.

It is worth noting that these cities have a lot of light industry and small manufacturing as well as a lot of people expert in electronics, programming and related fields. I have absolutely no doubt that some of these folks are taking a look at what Ukraine has done with drones and other improvised weapons like petrol bombs and considered how to adopt them to the local terrain – both dense urban and rugged rural. From advanced drone delivered improvised antitank weapons to low tech caltrops, the options for guerilla weapons that can be made by motivated groups with access to modern machine tools are broad and easily distributed and copied using the internet to hundreds of other similarly equipped little workshops.

I mean yes, the PRC could conquer Taipei. If you have a million men to spare – and technically you probably do if you are the PRC – and a way to get them onto the beachhead in large numbers fairly quickly with lots of tanks etc. But unless you know you have air superiority and good naval logistics, any force that you send via Taoyuan to Taipei is going to make the Russian advance on Kyiv look like a tactical success with low loss of life. And even if you do have both you are very much relying on the defenders to not be resupplied with munitions, which is a whole other problem

Logistics and Resupply

In addition to establishing your own naval access route, you have to blockade the island so that it doesn’t get resupply via Okinawa, which isn’t very far away (the closest island with a runway is Ishigaki ~150 miles away – the red circle in the first map above). The closest Japanese territory to Taiwan is about half as far which means that Japanese territorial waters extend to about 30 miles/50km from the Taiwanese coast at the closest point. Without an effective blockade resupply via Ishigaki or even the main Okinawan islands is very easy. In fact if there’s no blockade and a lack of effective air superiority by the invaders then resupply is going to use the Keelung container port and all the logistics handling there (see picture earlier). This is going to be massively more efficient that resupply via an improvised beachside dock. Just imagine how many artillery shells and missiles the Ever Given could transport if it were reasonably sure of not being attacked and then those containers full of munitions could be loaded onto trains and trucks and transported to any part of the island not under PRC control.

If there is a blockade then resupply is harder, but still probably not very difficult. Even assuming the US Navy decides to not get involved directly, and the Japanese decline to do more initially than provide a base for local resupply, hypothetical blockade runners are only going to need to go 30 miles (2 hours travel time for a fishing boat, far far less for a cigarette boat or other similar power boat) outside the defensive umbrella of the JMSDF.  It is true that the power boat resupply will not provide large amounts of weapons but it could provide a few tons of, say, NLAW or Javelin ATGMs. A medium sized fishing boat could transport significantly more, albeit at a greater risk of interception. But again the interception can only be in Taiwanese/Chinese waters and that is, as mentioned above just 30 miles from the coast. If the Taiwanese have land based anti-air or anti-ship missiles that they can use to protect blockade runners then quite a lot of the run outside Japanese waters will be protected by those land based assets, possibly the entire 30 mile distance will be depending on the assets and the nature of the blockading force. This is important because the PRC really really does not want to cause an armed incident with Japan while it is invading Taiwan that causes Japan to intervene militarily.

It is hard to say how Japan’s public opinion and politicians would react to an invasion of Taiwan, they might decide to trash the defence only bit of the constitution and overtly supprot Taiwan from the outset or they might not. I would expect them to both send their own arms to Taiwan as well as facilitate resupply. I am, howver, absolutely positive that if the PLAN or PLAAF invade Japan’s sea or airspace as part of the invasion or in pursuit of blockade runners the Japanese SDF will get involved in stopping them permanently. Moreover I strongly suspect that even if opinion is not in favor of actively assisting Taiwan before an incursion into Japanese claimed territory, it will be strongly behind getting involved if an incursion happens. The PLA really really doesn’t want to have the region’s second largest military (after the PLA) involved in a war against them.

Given the terrible performance of Russian troops and equipment in Ukraine, there is no doubt that many nations are evaluating the likely actual performance of their own military assets and those of their likely antagonists. I don’t think anyone is giving the Japanese SDF a lower grade than before. The JMSDF is the third largest navy in Asia (the world?) and it has equiment that is comparable to the US Navy – in fact it has some equipment (Aegis destroyers) that are effectively identical to the USN. Moreover it exercises with the USN and other allied navies and regularly demonstrates that it hasn’t forgotten the traditions and institutional knowledge of its parent the IJN in terms of systemology. For example, no one doubts that the two aircraft carriers helicopter destroyers Izumo and Kaga can sustain numerous sorties under fire if need be, on the other hand there’s a lot of scepticism about whether the PLAN can operate it’s aircraft carrier at all, let alone in combat. If the PLAN or PLAAF cause the JSDF to decide to take a more active role then the logistics that will be blockaded is the resupply of PLA forces in Taiwan. That leads to the nightmare where the PLA invasion force is left with no option but surrender because it is out of bullets and can’t get any resupply, unlike the Taiwanese.

And at that point we’re looking at the PRC nuking Taipei, the PRC then discovering that the Taiwanese had a sleeper cell that blows up the 3 gorges dam and then the likely start of WW3… But all that escalation is way beyond this blog post. I’ll just add a final point. All this is predicated on the Taiwanese deciding to resist like Ukraine and not surrendering en masse in the face of a PRC invasion. Recent polls have 70-75% of the population of favor of resistance which seems to be borne out by the votes in recent elections and the positions of the current elected leadership so that seems entirely plausible.