随着先进人工智能的出现，现在还能把舰队隐藏在海洋中吗？

Few-Sheepherder-1655
I doubt this has been feasible for decades. Any nation interested in tracking a fleet has radar satellites. If you know when they leave there is a circular error of probability that is greatly smaller than the ocean itself.

我怀疑这种做法几十年来都不可行。任何有兴趣追踪舰队的国家都有雷达卫星。如果你知道它们何时出发，其圆概率误差会远小于海洋本身。

OHHHHHSAYCANYOUSEEE
What do you mean? It has been feasible for decades?

你是什么意思？这种做法几十年来不是一直都很可行吗？

Jpandluckydog
Feasible, but in the “there’s a surface group in this general area of ocean” way, not the “here is the exact position of all the ships and their speed and bearing” way.

是可行，但那是在“这片海域某个大致范围里有个水面编队”的层面上，而不是“这里精确显示了所有船只的位置、航速和航向”的层面上。

OHHHHHSAYCANYOUSEEE
That has been feasible since forever. I’m asking about precise locations.

这个想法自古以来就可行。我是在问具体的位置。

Jpandluckydog
It hasn’t been feasible since forever, in fact it’s only been possible in the last few decades. 

事实上，这个想法并非自古以来就可行，而是近几十年才变得可能。

Satellites can’t give persistent firing solutions, no. It’s questionable whether you could even make them able to without running up against physical limits. 

不，卫星无法提供持续的火力解决方案。即使试图突破物理限制，能否使其具备这种能力也值得怀疑。

Significant_Key_2888
No. It's not even possible to hide air defenses, artillery, tanks, etc. We live in a kind of pre-WW1 environment where the experts have not caught up to the reality that distinct, easily classifiable geometries seen from space are not survivable. The US and China have recently taken to massively increasing their space-based assets since machine learning has made tracking from space a more active and perhaps decisive part of warfare.

不。现在甚至连防空系统、火炮、坦克等都难以隐藏。我们正处于一种类似一战前的环境中，专家们尚未认识到一个现实：从太空望去，那些形状分明、易于分类的几何体在战场上已经无法幸存。美国和**中国**最近开始大规模增加天基资产，因为机器学习技术使得从太空进行跟踪成为战争中更为活跃、甚至是决定性的组成部分。

OHHHHHSAYCANYOUSEEE
What year did the change take place? Is it recent or has it been like this for 10+ years?

变革发生在哪一年？是近期的变化还是已经持续了十年以上？

Significant_Key_2888
It became public knowledge during the war in Ukraine that the US had these capabilities. It was likely not possible 10 years ago.

乌克兰战争期间，美国拥有这些能力已成为公开信息。这在十年前可能还无法实现。

OHHHHHSAYCANYOUSEEE 
You mean their feeding of artillery targets to the Ukrainians?

你是指他们向乌克兰人提供火炮目标吗？

I still feel that is simpler than finding ships in the ocean but I could totally be wrong.

我仍然觉得这比在海洋中寻找舰船要简单，但我可能完全错了。

Significant_Key_2888
Ships are gigantic, they have clear contrast against the ocean and have enormous wakes. Military vessels also have very unique geometries that distinguish them from civilian vessels. If you can classify the specific variant of a much smaller vehicle from space against attempts at concealment, ships are a much easier problem.

舰船体积庞大，在海洋中形成鲜明对比，且会留下巨大的航迹。军用船只的几何外形尤其独特，与民用船只明显不同。如果能在太空环境下识别出经过伪装的小型车辆具体型号，那么辨识舰船显然是个简单得多的问题。

A big problem today is the credibility of nuclear deterrents of smaller powers. You can't rely on mobile launchers. Few countries have enough submarines to really count on that. ICBM fields with enough mechanisms to stay survivable and dummy silos cost a lot of money. Air bases are not survivable enough to first strikes. This is why China is investing massively in their nuclear forces now.

当前一大难题在于较小国家核威慑力量的可信度。移动发射装置并不可靠，而拥有足够潜艇数量以形成有效威慑的国家寥寥无几。具备高生存能力的洲际弹道导弹发射场与伪装发射井的建设成本极其高昂，空军基地在首轮打击中的生存能力也明显不足。这正是中国当前大规模投资核力量建设的原因所在。

Jpandluckydog
A big problem in that argument is that China has been building exclusively silos and road mobile TELs.

这一论点中存在的一个大问题是，中国一直在专门建造地下发射井和公路机动运输起竖发射装置。

GreatAlmonds 
They've also been investing very heavily in their SSBN based deterrence.

他们还在大力投资基于战略核潜艇的威慑力量。

The JL3 was revealed last year, giving the existing type 094s the ability to hit continental USA from Chinese waters.

去年亮相的巨浪-3 导弹让现有的 094 型战略核潜艇具备从中国海域打击美国大陆的能力。

The next generation 096 SSBNs are also in development / production.

新一代的 096 型战略核潜艇也正在研发或生产之中。

Jpandluckydog
“Very heavily” is a stretch. Not an expert China watcher but it seems pretty clear that it’s been plaRF->Surface Ships->Munition Inventory->ISR->Submarines, in terms of budget priority. You can mix the stuff in the middle around but the ends are correct. 

"大力"有些言过其实。我虽非中国问题专家，但预算优先顺序明显是火箭军→水面舰艇→弹药库存→情报监视侦察系统→潜艇。中间项目的顺序或许可以调整，但首尾两端是正确的。

What has actually received heavy investment is their missile forces, which have genuine world-beaters in their arsenal. That’s what the results of very heavy investment looks like for China, and their subs are very far from that, and still much worse than pretty much every other SSBN armed nation. 

真正获得巨额投资的是他们的导弹部队，其武库中确实拥有世界顶尖的装备。这才是中国重金投入的成果体现，而他们的潜艇远未达到这种水平，仍然明显落后于几乎所有其他装备弹道导弹潜艇的国家。

However, I do think they will accelerate it in the coming years, as a response to American investment in CPS and CPS-like capabilities. The ICBM investment was a response to US missile defence more than anything. 

不过，我认为他们未来几年会加速发展潜艇力量，以应对美国在常规快速打击系统及类似能力上的投入。此前中国对洲际弹道导弹的投资，主要是针对美国导弹防御系统的回应。

wrosecrans
Exact capabilities of spy satellite stuff is very classified, so... shrug. Hard to say.

间谍卫星的确切能力信息属于高度机密，所以……只能说很难判断。

But yeah, I think if you were willing to throw a bunch of money and classified technology at it 10+ years ago, it'd be surprisingly possible to keep track of something like a CSG. Finding one from scratch is hard. But if you know where it starts then you "just" have to not lose it. If you don't have a starting point, then you need very high res images to say "this specific speck is a carrier, that speck is a cargo ship." But with low res imagery that you can get much wider/faster coverage with, you can say "there's a 10 km long wake here, and there's probably a boat at the front of it that is too small to see directly in this image" and correlate that with the previous 1000's of images to say that it is the boat that left XYZ naval base two weeks ago.

不过确实，我认为如果十多年前有人愿意投入大量资金和机密技术，追踪像常规动力潜艇（CSG）这样的目标其实是出人意料的可行。从头开始发现一艘潜艇很难。但如果你知道它的起始位置，那就"只"需要别跟丢。如果没有起点，那就需要极高分辨率的图像来断定"这个特定的点是航母，那个点是货船"。而使用分辨率较低的图像虽然细节不足，但能获得更广、更快的覆盖范围，你可以判断"这里有一条 10 公里长的尾迹，前端可能有一艘船，但在这张图像中太小无法直接看到"，然后与之前的数千张图像进行比对，从而推断出这就是两周前离开 XYZ 海军基地的那艘船。

Did the US actually invest the R&D to make that a practical capability? I have no idea, and a lot of our doctrine has a lot of inertia dating to the cold war. Was it at least theoretically possible? My gut says yes it was probably possible. But until quite recently, there wasn't an adversary carrier fleet that we were terribly worried about possibly losing track of, so my best guess about in-practice is that way more classified R&D money was going to finding subs than tracks large surface ships. And overhead optical tracking sounds like a terrible way to track a deeply submerged submarine.

美国是否真的投入了研发来使这项技术成为实用能力？我对此并不清楚，而且我们的很多军事理论还深深植根于冷战时期，具有很大的惯性。至少在理论上是否可行？我的直觉是，很可能确实可行。但直到最近，我们还没有一个真正令人担忧的、可能会失去踪迹的敌对航母舰队，所以，对于实际情况，我最好的猜测是：在投入大量保密研发资金的项目中，追踪潜艇的比重远远超过追踪大型水面舰艇。而使用高空光学追踪来追踪深潜潜艇，听起来确实是个糟糕的主意。

jellobowlshifter
An answer from the perspective of the US looking for Chinese carriers? How bizarre.

从美国搜寻中国航母的角度来回答这个问题？这未免太过奇怪了。

wrosecrans
Well, we've got more spy satellites. So we'd be more likely to be able to track a carrier group than anybody else. If asking if it's possible, considering the largest satellite constellation seems like the most obvious place to start musing about the possibility.

嗯，我们拥有更多的间谍卫星。因此，与其他任何国家相比，我们追踪航母编队的可能性都更高。如果询问这是否可行，考虑到最大的卫星星座似乎是最显而易见的出发点，来探讨这种可能性。

jellobowlshifter
Well, there still isn't an adversary carrier fleet that you're terribly worried about possibly losing track of.

话说回来，目前仍然没有一支需要您极为担忧、可能会失去踪迹的敌对航母舰队。

ActionsConsequences9  
In the desert sure, but not in concealment under trees urban areas et al, Israel had US intelligence and still could not stop the missile launches (but it was much better than the Persian Gulf war).

在沙漠地带确实如此，但在树林、城市区域等遮蔽环境下则不同。以色列虽然拥有美国的情报支持，却依然无法阻止导弹发射（不过情况已经比海湾战争时好得多）。

helloWHATSUP
cloud cover   

云层覆盖

most modern satellite uses synthetic aperture radar, so clouds are not a factor.

现代卫星多采用合成孔径雷达，因此云层不再构成影响。

But yeah, modern satellites will spot and auto identify anything as obvious as a carrier within seconds. And if it's within ~1000km of the Chinese coast they can just send up a drone to track the carrier in real time with radar from 600? km away.

确实，现代卫星能在数秒内发现并自动识别如航母这般明显的目标。若其位于中国海岸线约 1000 公里范围内，中方甚至可派遣无人机在约 600 公里外通过雷达实时追踪航母动态。

ActionsConsequences9 
I wonder what is the state of EW vs SAR from sats.

我想知道从卫星上看电子战对合成孔径雷达的效果如何。

Jpandluckydog
Well, ship based jammers can push out much, much, more power than the satellites themselves. If conventional wisdom is true, both the traditional and “advanced” kinds of jamming should work very well, alongside the optical dazzlers that I’m pretty sure are already on USN ships and maybe one European ship. 

船载干扰系统输出的功率比卫星本身强得多。如果传统观点成立，无论是传统还是“先进”类型的干扰手段都应该效果显著，更不用说我很确定美国海军舰艇和可能某艘欧洲舰船上已经装备的光学致盲器了。

Far from an expert and atmospheric conditions can do a lot of weird things to radar waves, so take this with a grain of salt. 

我远非专家，而且大气条件会对雷达波产生许多奇特影响，所以请对我的看法持保留态度。

ActionsConsequences9  
Ultimately it is the nature of the echo that has been the forefront of EW jamming, visible spectrum is high energy with limited echo so you need a super intense energy source to blind, not even the sun is powerful enough based on the digital camera footage of it. You see something very bright in the corner but you can see around it easily enough.

归根结底，回波特性一直是电子战干扰技术的核心问题。可见光频段能量虽高，但其回波有限，因此需要极强的能量源才能实现致盲效果——即便是太阳的强度也不足以达成此目标，这一点从数码相机拍摄的日食影像中便可得到印证：在画面角落能看到极其明亮的光斑，但周边景物依然清晰可辨。

Other frequencies can absolutely blind radar without that much power, so I do kinda wonder how jamming would work on SAR, would it be blips of noise in the ocean hiding with hundreds of dots of noise or do they completely make the satellite completely useless while pointed at it? It just has to be the former.

而其他频段的电磁波完全能够以较低功率实现雷达致盲，这不禁让人思索合成孔径雷达系统面临干扰时的工作状态：究竟是像海面上零星散布的噪点般呈现为脉冲式干扰，还是能让对准干扰源的卫星彻底失效？答案必然是前者。

Jpandluckydog
Visible spectrum satellites are the realm of optical dazzlers (lasers). They’re new, though, and only the US and to a limited extent some European nations have really been investing in them. And to clarify, they’re much brighter than the sun. 

可见光卫星属于光学干扰器（激光器）的范畴。它们虽然新颖，但目前仅有美国以及部分欧洲国家在有限范围内进行投入。需要明确的是，它们的亮度远超太阳。

Radar jammers are a whole different system. I’m not going to even speculate as to what it actually looks like on the screen, but keep in mind with modern jammers there’s a lot more techniques than dilution/noise jamming like you’re describing, like attempting to make false positives. 

雷达干扰器则属于完全不同的系统。我不打算推测它在屏幕上实际呈现的效果，但要记住，现代干扰器除了你提到的稀释/噪声干扰外，还有更多技术手段，比如尝试制造误报。

ActionsConsequences9 
Good point on lasers, never considered a coherent beam for some reason.

关于激光器这一点说得很好，不知为何我从未考虑过相干光束。

SericaClan
The end of hiding fleets in the ocean will not be due to some fancy words like AI, but the advent of big LEO satellites constellation now enabled by reusable rockets.

舰队无法在海洋中隐藏行踪，并非因为 AI 这类时髦词汇，而是可重复使用的火箭使大规模低轨卫星星座的部署成为现实。