“Homeland Cruise Missile Defense” Panel 1 – Report Rollout: North America is a Region, Too

Available Downloads

Lee Hudson: Thanks for having me today. My name is Lee Hudson, defense reporter at Politico.

Also, joining me today are Dr. Tom Karako, the director of the Missile Defense Project and senior fellow at the – here at CSIS at the International Security Program. Also joining me today are Colonel Matthew Strohmeyer, military fellow at CSIS; and Wes Rumbaugh, the associate fellow with the Missile Defense Project at CSIS.

Each of our panelists are going to give remarks, and then we’ll have time for discussion.

Tom Karako: OK. Well, thank you, Lee, and thank you, Seth. And also, thanks to the deputy commander, General Roper, for that great introduction and that statement of support.

You know, Seth quoted Admiral Winnefeld’s remarks back from 2015 – it is kind of hard to realize it has been that long – when we launched the project, and a recurring feature of our work since has been an emphasis on, you know, the different aspects of air and missile defense besides just the legacy rogue-state ballistic missile problem, and especially the need to redirect to the near peer.

Today’s report channels some similar themes. It has been a big undertaking. I want to thank the whole Missile Defense team – Ian Williams, Wes Rumbaugh, Shaan Shaikh, and Masao Dahlgren, as well – as Lee said, Colonel Matt Strohmeyer, our Air Force military fellow.

We’ve also benefitted from a lot of other folks, including our co-author and advisory board member Ken Harmon, who’s here today, and a lot of other SME and outside reviewers.

Let’s begin by going back to Admiral Winnefeld’s comments that Seth referenced where he said the homeland cruise missile defense had risen in importance relative to regional BMD. It was just after the invasion of Ukraine, and various Chinese shenanigans in the South China Sea. It was also just before the unfortunate JLENS incident, which was an experimental aerostat nearby here designed to get after just this kind of problem.

Admiral Winnefeld posed the most important, I would say, threshold question for today, which is why. Why is it that we should defend against cruise missiles or UAVs, for that matter, if we’re not going to contend with the big nuclear ICBM threat from Russia and China?

For years, many folks have answered this question in the negative; that we don’t need to. After all, the Soviets had nuclear SLCMs, and we chose to deter them and other forms of nuclear attack with the threat of retaliation. We’ve got nukes; they wouldn’t dare. They’ll be deterred QED.

I would say that unimaginative formulation is almost – it seems like it’s based on a kind of hubris about the homeland as a sanctuary. It’s unfortunately too representative of conventional wisdom today. The problem we’re facing is not a lesser-included set of strategic nuclear attack. Rather, non-nuclear attack with strategic effect, whether by kinetic means, or non-kinetic, or both, is a different problem about what an adversary might think they can get away with beneath the nuclear threshold.

Unfortunately that problem is worsened by how U.S. air and missile defense efforts have long been characterized by a dichotomy. Defenses for the homeland have focused on long-range ballistic threats while other air defense efforts have been limited to regional and force protection applications, almost to the exclusion of the homeland. And that compartmentalization assumes that battles in one place will only consist of certain types of threats, and battles elsewhere will consist of others. But a changed environment, a strategic environment and the proliferation of sophisticated threats have come to make that dichotomy increasingly obsolete. Whether we used to call it national and theater, or homeland and regional, it kind of ignores the basic, simple fact that North America is a region, too, and as such, the full threat, the full panoply of UAV cruise missile and other threats that we have to be concerned about over there may also matter here closer to home.

While the 2019 Missile Defense Review well described the missile threat spectrum, neither its actions nor the associated budgets and programs at that time did that much about it. Beyond taking note of studies underway and promising to designate an executive agent, we’re mostly left to admiring the problem. And as a result, time has ticked by. The near-complete lack of homeland cruise missile and related forms of air defense has created a deterrence problem and a vulnerability that near-peer adversaries now seek to exploit. An adversary seeking to change America’s strategic calculus may be tempted to employ long-range, conventionally armed strikes to achieve strategic effects while remaining beneath the nuclear threshold. Deterring this kind of attack requires an element, at least, of deterrence by denial.

Now instead of thinking about missile threats as some kind of boutique threat, we should learn the lessons of, frankly, the last decade or so, and recognize that they are – in the words of Assistant Secretary of Defense John Plumb – weapons of choice. When I hear folks talking about truck bombs as a more desirable way to attack the homeland, I feel like I’m in a time warp. It’s all very 1995.

But just in case you were under a rock, we’ve seen tactical missiles, and cruise missiles in particular, used in large numbers – 2800 and counting in the first 125 days of the Ukraine war. But of course there was also Nagorno-Karabakh; the Abqaiq attacks of 2019 in Saudi Arabia where President Biden is visiting this week.

The center of this missile threat spectrum is the humble but reliable cruise missile. Hypersonic this, that, and the other thing gets a lot of the buzz, but it’s the sturdy land attack, anti-ship cruise missiles that everyone seems to reach for early and often.

We see here on the right the Russian Kalibr sea-launched cruise missile and the kind of stand-off range it might have. And on the left, the air-launched Kh-101 or AS-23A. Now the Russians claim that the Kh-101 has a range of some 4500 kilometers. This graphic, from a polar perspective, shows what that reach gets you, even if fired from a stand-off position above the north warning system. When Russia fired Kalibrs into Syria from thousands of kilometers away in 2015, of course, they didn’t do so because they were concerned about the A2/AD bubbles from ISIS. It was – it sure looked like messaging about their capabilities.

Following Seth’s thanks to the combatant commands, I should single out the leadership of one other person here, General VanHerck, who has – probably more than anyone – been responsible for shaping the conversation about the strategic problem we are discussing today. This is a quote from his congressional testimony in March I believe it was, highlighting that vulnerability and deterrence gap beneath the nuclear threshold.

But again, it’s a mistake to think of this as a NORTHCOM idiosyncrasy. A threat to North America is about the relationship to other geographic and functional combatant commands, and thus America’s role in the world.

So I commend to you an op-ed from May 31 by both General VanHerck and General Jacqueline Van Ovost, the head of USTRANSCOM. Why TRANSCOM? Some 85 percent of the joint forces within the U.S. homeland, in an effort to shape calculations, could include efforts to impede our flow of forces and power projection. General Roper this morning said, to bring us to our knees.

The old-think legacy of that homeland regional dichotomy and its manifestations in our UCP, our organizational structure and resourcing, and our approach to air and missile defense has created gaps and seams that our enemies will exploit. Which brings us to the need to thwart or blunt that threat and introduce doubt that it would succeed.

The Biden administration is to be commended for not adopting a no-first-use nuclear policy, but deterrence by the threat of punishment is not enough. It will have to be a mix – a mix of deterrence by denial as well.

We want to give credit to the Congressional Budget Office for releasing a report just over a year ago on this topic. They were tasked to look at architectures and elements that could contribute to nationwide cruise missile defense. It helped advance the discussion, it has a lot of really useful budget data. It was a very instructive contribution.

Unfortunately, their architectures were a bit impractical and, frankly, unaffordable. A few assumptions deserve particular attention: First, an apparent focus on defending everything, an apparent conflation of the defended asset list with the critical asset list; second – closely related – a perimeter-based defense design which lacks defense in depth and layers, and only one type of sensor per architecture, with elements – while selected to maximize detection range and decision time – which is understandable – it also impaired the quality of the defense; and finally, a relative over-reliance upon fighters, especially for identification and engagement. Again, an understandable choice because of the sensitivity of shooting at things in American air space. But the combat identification challenge, while real, is not necessarily best solved by having a fighter pilot try to find an incoming missile and look out the window to positively confirm it’s a threat. Colonel Strohmeyer is a fighter pilot so he can tell you why that might not be ideal.

This is our image – not theirs – depicting the cross-range challenge of even getting fighters there on time. But again, it’s an important report, and I commend it to your reading.

So we decided to take our own crack at the problem. Now I know what you’re thinking: boy, oh, boy, a think tank report waving generally in the direction of some desirable capability – I know, I get tired of those, too. So we decided to do something a little different: building a defense design and costing it. But first we identified some principles to inform that design.

The beginning of wisdom is that you – as we heard from General Roper this morning – is that you can’t defend everything; it would be foolhardy to try. Unlike the ballistic missile discussion of the 1990s, frankly, population centers may not necessarily be the most significant target for non-nuclear strategic attack. In this report, we don’t go to the opposite extreme, however, of merely point defense, and instead point to a broad-area-defense approach, which is – turns out is pretty useful because there is a lot of critical assets that are clumped together in certain parts of the country.

Second is multi-mission. When asked about this topic recently, Deputy Secretary of Defense Dr. Kathleen Hicks emphasized the need to approach it in an integrated and multi-mission way. And you’ll see in our architecture the toothing for adjacent missions like counter-UAS, and more advanced problems like hypersonic threats.

Third, the full-attack life cycle – not merely earlier indications of warning, but also means to hold at risk bombers or submarines. As General VanHerck and Admiral Richard have talked about, influencing the threat and communicating to the enemy – campaigning as it were – to pass along that we’re on to them.

Defense in depth – in a way, the opposite of a perimeter defense or point defense with an enhanced warning time and multiple opportunities for threat classification and decision time, but also layers of sensors and effectors.

Fifth, it’s nice to say that we want to maximize mobility and flexibility, but sometimes the best can be the enemy of the good, both for cost reasons and also to resist the temptation to fly things away to other parts of the world at the drop of a hat. It may be better to have some fixed and dedicated assets that are bolted down, dedicated to the homeland which is, after all, our top defense goal.

Second – or sixth, throw nothing away. And here we have a special opportunity, I’d suggest, to test the department’s concept of integrated deterrence, which is to say pulling together both defense and non-defense assets. There’s a lot of things within CONUS that can and should be brought together. Many of them are not DOD.

And finally, affordability. Even if on paper homeland defense is the top goal, we all know that it will not be pursued if it’s going to cost a half trillion dollars. And this principle is a measure of the success of all the prior ones. To this I would add a corollary, and that is of timeliness, and it needs to be relevant to the threat. This means substantial activities in this decade, not waiting until the next. It means we cannot wait for the completion of exotic space-based central layers nor for the nirvana of JADC2, should it eventually be realized.

In a few minutes you are going to see the elements, layers, and phases we lay out, and our purpose in doing so is not to say that this is the end-all, be-all, or the precise solution. On the contrary. Our point here is to show that this is a soluble problem, and it does not need to cost a half trillion dollars.

Since we’re pretty critical of everybody else, I’ll be self-critical. Instead of trying to defend everything with the result that nothing is defended especially well, we kind of do the opposite, and we go to a very thick defense of a handful of areas with multiple layers of sensors and interceptors. In other words, we went high – high-end on capability.

So what you are going to see in some respects may be too robust, even if it is a fraction of the cost of other folks. And as you will hear from Wes, we also went very conservative on the costing estimates at every turn.

The elements we used as the baselines – consistently at the higher end and multiple types. We also added cost categories that are not present in CBO for things like systems integration. So it’s a robust architecture, but there’s lot of room for paring it back a bit for still-lower cost solutions.

As the deputy secretary has said about this issue, our thinking here has to change. It’s not just about rogue-state ballistic missiles anymore. The threat gets a vote, it’s already voted, and it has voted for a spectrum of especially aerial threats. And the sophistication and the kind of assumptions that we have long brought to bear in the regional contexts must now be applied here at home as well.

The good news is that efforts to realize homeland cruise missile defense have already begun. The road to homeland cruise missile defense for North America will probably go through Guam, and efforts for the defense of Guam will be especially instructive for element selection, systems integration, and command and control development. Both PB-22 and 23 included funds for homeland cruise missile defense – experiments, towers, OTHRs – both of which you will see in our architecture in spades.

Vice Admiral Jon Hill of the Missile Defense Agency has said that the defense of Guam is probably the hardest problem his agency faces. It may also be the most important. If we can get that right – and we will – its lessons will inform global air and missile defense efforts suited, after all, to the central challenge of our time.

So to walk through our defense design, let me turn things over to Colonel Matt Strohmeyer – call sign Nomad. Nomad is a career Air Force pilot and operational war planner, and a strategist who has been the genesis for defense concepts like agile combat employment and strategic shaping. He was most recently at USNORTHCOM and NORAD where he designed and led General VanHerck’s global information dominance experiments.

Matt, over to you.

Colonel Matt Strohmeyer: Well, thank you very much, Tom, for that kind introduction. And thank you for the other panelists here – for Lee, for moderating; for Wes and the other authors on the report, and also for USNORTHCOM, and General Roper, and Seth for their remarks this morning.

So I think it’s important to note that while I’m an Air Force fellow at CSIS, my views and those expressed in our report do not necessarily reflect those of the Air Force or the Department of Defense. And with those caveats out of the way, let’s just dive into our homeland defense design.

So in designing a capable defense for credible deterrence, we sought to apply the seven principles that Tom just highlighted, and then apply them across what becomes for us five different layers. You can think of these layers as essentially our defense in depth; rather than having one brittle layer, having defense across the life cycle of a cruise missile or other types of threats. And those layers that we’ll walk through are resourced with individual elements. You can think of these as the capabilities, the technologies that go inside each one of those layers. And then the report has those layers and elements phased in over the course of three phases for the next eight years.

So to start with the first layer – layered defense in depth – what does that look like? For us, defense in depth and the five layers are all about time and options; increasing the amount of time available to influence a threat.

And our design starts far left of launch of that cruise missile threat, and even left of a bomber or submarine threat that it might emerge from, and going left as far as getting to the decision makers that will decide to attack North America. This first layer – global threat awareness – seeks to apply global rather than regional perspectives to how we see strategic competitors and how they might take offensive action against us, allowing us to take more of a global response to deter those actions. This layer is all about gaining time and decision advantage. It seeks to go far left of launch, the days and even weeks prior to a possible attack to identify changes in adversary patterns of life, and then to identify proactive options for how we might deter such an attack. The layer builds on recent efforts by NORTHCOM and NORAD and their GIDE – or Global Information Dominance Experiments – to move from a more regional – depicted here – to a more global perspective.

In the days leading up to a potential attack, competitors may change their pattern of life in ways that may be observable, especially if we are sensing on a more global perspective as they look to set their force for an attack. The advent of increased commercial space sensing and the application of artificial intelligence to these sources can allow, in the aggregate, for us to gain a better understanding of changes in those adversary patterns of life. These insights can then give opportunities to combatant commands to rapidly collaborate and take equally global actions to deter. These type of actions may be an example of what the department refers to as integrated deterrence.

The second layer for us – building on that global threat awareness – is what we refer to as the 21st century version of the Distant Early Warning Line – or DEW Line – of the Cold War that provides 360-degree early warning of air and maritime threats to North America. The goal of this layer is not a high-fidelity tracking and identification of threats, but rather just initial detection and then custody of those threats.

The technology that we are leveraging for this layer is known as over-the-horizon radar. Over-the-horizon radar – or OTHR – uses large arrays of radars that – radar signals that are bounced off of the ionosphere to detect air and maritime targets at extreme ranges of over thousands of kilometers. The goal is not, again, high-fidelity tracking of threats, but rather the ability to search large areas and identify and maintain custody of those threats, providing not just minutes, but hours of early warning that can increase our decision space and the response options that General Roper talked about. OTHR is a proven technology. It’s ready today and leveraged by Australia through their Jindalee over-the-horizon radar network – or JORN – and was recently announced by Canada as having a major investment in OTHR for warning and awareness of the northern approaches to North America.

Inside that second layer is our third layer, known for us as wide-area surveillance. This layer seeks to maintain custody of possible threats coming from those initial OTHR detections, allowing us to refine possible target areas and then prepare defense options. This layer combines existing active and passive sensors, many of which were not originally designed to track missile threats, for an integrating sensing network to maintain low-fidelity custody as threats approach those target areas. Sensors that are – that would be integrated into this layer include things like FAA’s air-traffic radars. These radars and the algorithms on them were designed for medium- and high-altitude air traffic, and have greatly reduced coverage – as seen here with the green and yellow indications – for low-altitude objects.

NORAD recently demonstrated the ability to take the existing radars and unlock the raw data coming off them and identify targets not seen with those legacy algorithms. And while these sensors are planned for upgrades, maintaining custody of threats at low altitude over North America requires the integration of additional active and passive sensors – active sensors such as NOAA’s weather radar constellation that exists across North America. These weather radars, which are depicted in here, can be integrated to augment the detection capability. Sensors such as these radars cover much of North America and, while they have a slow scan rate and limited low-altitude ability, could provide another source of custody if integrated with other existing sensors. You can see here depicted one of those NOAA weather radars that were able to pick up a flock of birds as it took off, in this case those birds being very low radar cross-section targets, indicating that it is possible to use those radars for tracking air targets.

We also need to be able to integrate other sensors such as, for example, Customs and Border Patrol’s constellation of aerostats, or TARS, that exists along the southern border of the United States and that are also able to sense air targets out to 200 nautical miles. Integrated into a larger homeland defense system, these sensors could supplement custody of threats that are in transit over North America.

But it’s not just active sensors; it’s also passive sensing that can help maintain custody for – just like back in World War I, the passive sensing seen here could help us maintain warning and awareness of those air threats. If we know – just the location and power of an emitter such as cellphone towers or radio or TV tower emitters can allow us to get a rough geolocation of noncooperative air threats through passive and multistatic sensing. North America, as seen here, possesses a very robust constellation of those sensors that can allow us to be able to bring that passive sensing together, combine it with active sensing to get a more integrated picture for wide-area surveillance.

The fourth layer for us is one of the most – and one of the key layers as far as our preferential defense – is something that is known as PADs, or prioritized area defenses. Rather than defending all of North America, these PADs provide robust detection and defense against low-altitude missile threats over critical targets in large defended areas. These PADs consist of networks of sensor towers that are combined with persistent surface-to-air interceptors to provide a highly capable defense of the top tier of targets against North America.

So, why a PAD of networked sensor towers instead of a single high-power sensor? Well, it turns out that the Earth curves, and that curvature creates a real challenge for detecting low-altitude targets at range. The radar horizon, for example, for a ground-based sensor regardless of its power is approximately 40 kilometers, allowing a low-altitude cruise-missile threat to sneak underneath that sensor to the target area. Elevating a sensor, while it can help mitigate the radar horizon challenge, comes with its own challenges of getting enough power to the radar to allow it to see the threats at the ranges that you need to be able to expect an intercept.

By contrast, in our design, by leveraging lower-cost towered radars seen here in yellow and EO/IR sensors on those towers, a PAD can cover a larger area while simultaneously maintaining robust detection and tracking of those low-altitude threats. A template PAD that’s seen here is made up of 19 networked sensor towers with both medium-range radars and EO/IR sensors to aid not only detection and tracking, but also identification of threats. These towers are combined, then, with medium-range interceptors located here at four notional locations that provide 48 different interceptor missiles to incoming threats. And those are combined within a long-range interceptor near the center of the PAD with eight missiles that allow us to be able to handle a range of threats potentially up at low altitude.

When we put this PAD together it covers over 500 kilometers in diameter, and that’s an area of around 200,000 square kilometers. Included in the PAD design is a multi-mission integrated test bed. Think of it is like local JADC2 to allow for integration of emerging sensors and affecters, as well as the command and control of those systems. This test bed and the associated operators would be federated up to the national homeland defense level.

So the use of a PAD design with overlapping sensor coverage is all about time. A threat that’s approaching the PAD at low altitude from thousands of kilometers away would initially have been detected by over-the-horizon radar and then maintained custody of by the wide-area surveillance as it approached the edge of the PAD. Once over the PAD itself, the assumed target – with an assumed target at PAD center, the distance of 250 kilometers that that threat would transit over the PAD provides or equates really to what is referred to up here as time to decide – how much time defenders have to determine if and how to engage the threat. In the case of the templated PAD, defenders have almost 15 minutes of high-quality detection and tracking time over the PAD to decide to engage with enough time to keep that threat out of the specific PAD-defended area.

The fifth layer for us beyond the PADs themselves is risk-based mobile defense. So with these PADs that, obviously, provide the static persistent sensors and shooters, they’re augmented by this fifth layer. This layer includes fighter and airborne early warning assets like the E-7 Wedgetail to allow for both forward deployment to cover the launch platform avenues of approaches like adversary bombers over the Arctic and also to provide additional defense over areas that potentially are at higher risk given the specific threat or early warning detections. This can not only help us to channelize possible avenues of approach for adversaries, but also to create deterrent effects by holding at risk the launch platforms even before those cruise missiles would be released.

The various five layers of our defense design that we’ve outlined here are deployed over the course of three phases over the next five years – or, excuse me, the next eight years.

The first phase includes the deployments of the prototype PAD over certain critical areas that you can see here on the Eastern Seaboard. In addition to that initial deployment of the prototype PAD, we’d begin the integration of those wide-area passive and active sensors throughout much of North America. And then critically in this first phase is the deployment of four over-the-horizon radars facing the northern approaches of North America and Alaska.

The second phase deploys four additional over-the-horizon radars – two to the East Coast and then two to the West Coast – and then four additional PADs over several different critical areas over North America.

The third phase sees the expansion of several of the PAD capabilities, up to what is known to us as a PAD+. This PAD+ increases the existing PADs and the existing sensor towers up to 25, allowing it to see a(n) approximately 350-kilometer radius from left to right, depicted here. It also increases the number of medium-range interceptors to 72 and adds a multi-mission interceptor that can cover almost the entire radius or the entire area that the PAD covers. This allows us to start expanding our missions both – not just from conventional slower cruise missiles, but also moving potentially into hypersonic missile defense. That combined PAD area, as you can see here, for one of the locations – namely, the first PAD – also includes an aerostat that we would – that would be placed over that PAD for additional coverage.

That third phase, as depicted here, not only includes the addition of those PADs and expanding those PADs out, but also includes two over-the-horizon radars covering the southern approaches, which brings our total coverage to 360 degrees for that 21st-century distant early warning line. It also includes the addition of the ability for fighters to be able to rapidly forward deploy up to places like the Arctic to try to hold those avenues of approach at risk, and also the ability for us to use that episodic coverage from things like the E-7 to, again, restrict the avenues of approach of potential attacks. When we look at the coverage that the PAD alone provides and not even the distant early warning line or the wide-area surveillance, you can see that we gain a very robust coverage of low-altitude threats over most of the areas that those PADs cover.

And to complete PAD architecture, when it’s – when it’s put together it provides not only a capable preferential defense of North America, but also a credible deterrent to non-nuclear strategic attack. This defense design, however, is not built to be static, but to continually evolve and integrate emerging sensors, capabilities, and missions. These missions will undoubtedly include defense against hypersonic threats as well as slower small UAS threats. From a sensing perspective, space-based sensing of live threats will certainly be an important aspect not only for cruise-missile threats, but other emerging threats as well.

I’ll turn it over to Wes Rumbaugh, who will run through how this defense design not only provides that capable homeland defense but also one that is economically feasible in the near term. Wes, over to you.

Wes Rumbaugh: So thank you, Matt. And I’d just like to add my thanks to Tom, Ian, and Ken, our co-authors, as well as all of the reviewers who helped us with this report, and then also the rest of our team. And of course, thank you to Lee for moderating the panel today.

So I’m Wes Rumbaugh, and I’m an associate fellow here with the CSIS Missile Defense Project. And in particular, I do most of our work on the missile-related elements of the defense budget, so now it’s time to eat our vegetables. Don’t worry, though; there’s only one slide with numbers on it, so I tried to keep a lot of it sort of at the top level and examining the assumptions that go into the cost modeling. And then, if you’re curious for the gory details of all of the cost analysis, that’s why we have the appendix.

But at the outset, I think it’s useful to contrast kind of the cost of cruise missile defense and ballistic missile defense, which is probably what most informs people’s sort of generic mental cost baseline for how much things cost. Air defense is generally cheaper on a platform-to-platform basis, but it also requires more of those platforms to generate area coverage. So we’ve built a relatively robust and thick defense because we think that layering and defense in depth are important, but that’s also sort of a fundamental tradeoff in cruise-missile defense and ballistic-missile defense. You don’t have to invent a bunch of stuff the same way that we had to when we started deploying GMD, but you have to sort of shop at Costco for the capacity side.

So I want to start first by just comparing some of the assumptions of the CBO, cost estimates in particular. And some of this overlaps with Tom’s earlier discussion of the CBO study, but I want to focus on particularly how these things – how these assumptions affect the overall cost of the architectures.

So I think in the discussion of affordability of homeland cruise missile defense, most of the estimates or the disparity in estimates over how much these systems will cost is driven not by mathematical things or not by differences in data sources, but by the assumptions that are used to put together the calculations and how the architectures are put together. So I think that the first – the first part of the table, the primary cost drive is in defended area, because it determines how many assets you need for the defensive task.

And so the CBO’s assumption of needing to create a defensive perimeter around the United States creates significant requirements for number of platforms that drives their costs up. Whereas we have a preferential defense, you can sort of concentrate sensors and shooters in key areas while still requiring fewer total platforms, and thus reducing the overall cost. This, I think, is the key component of an affordable homeland cruise missile defense architecture.

The second assumption that the CBO – of the CBO study that increases the cost, I think, is the reliance on a single sensor type. It’s necessary in their perimeter-based defense for the warning times that they would need, but it forces the CBO to examine basically exclusively more expensive aircraft and satellites, exclusively. So the CBO rightly points to the significant benefits of putting sensors at higher altitudes for air defense, like we sort of talked about earlier, but there are also important limitations. And most importantly for the cost side, that decision precludes CBO from looking at towered and, more importantly, ground-based radars and their architectures, which offers some opportunities for cost savings.

This also kind of transitions to the third point, which is about sensor persistence in particular. So the way that they achieve persistence in their architecture with aircraft requires operating them 24/7, 365. And so flying aircraft for that long and that persistently gets expensive. It drives up your operations and sustainment costs fairly substantially. Our architecture, by comparison, uses sort of these towered radars and ground-based radars for the persistent element. And so that saves a little bit of cost.

I think the fourth thing that we looked at that’s a little bit different is the assumptions about system integration. So CBO study had a single sensor paired to basically a single interceptor site. And so they didn’t assess an additional cost for system integration. The diversity of platforms that we looked at, I think, required us to sort of have a more conservative estimate in regards to the cost of system integration. And so we added an additional $500 million on the initial portion of the architecture for those system integration efforts.

And then the final major difference is in terms of engagement method. CBO analyzed a mixture of fighter aircraft as well as long-range SAMs. And those additional fighter aircraft in particular can drive significant costs. Instead, we looked at a mixture of medium- and long-range interceptors which, again, contains costs. And we’ve also spoken to some of the operational reasons.

Among all those differences, though, I do want to emphasize how valuable the CBO study was, particularly as a data source, just in terms of its compilation of public data and its ability to publicize a lot of that data. We had to do something similar in terms of basing our estimates on public sources of data and finding analogous systems to be able to cost some of these things, and with some adjustments for changes in our configuration. Where the CBO had relevant data, we attempted to use it as much as possible, and generally used the most conservative estimates in their cost ranges.

So I promise, this is the only slide with numbers. But this is our sort of total assessed costs for each of the component systems, and then the totals for the CS architecture. So with the PAD and PAD-plus, all the way though the Phase III, and then sustainment over a 20-year period. So I think at the outset, some of you might be thinking that the sustainment numbers are a bit light. And that comes down to another methodological decision that we made, that I’ll talk about on the next slide. But first I want to talk through some of the decisions we made for the costing of each of the elements, and how we kind of derived some of those things. But there’s a more thorough discussion in the appendix of the report.

So for the over-the-horizon radars and the PAD radars, we lacked good, specific data for acquisition costs, specifically for the over-the-horizon radars because it’s sort of a new capability. So it’s something that’s still in RDT&E, so we don’t have unit costs. One of our subject matter experts gave us a total acquisition cost of $600 million, which we then divided between MilCon costs and the actual radar costs. And that was important because – so the most difficult thing for both the over-the-horizon radars and the PAD radars was calculating a sustainment figure.

The way that the Department of Defense usually reports operations and maintenance and sustainment data is based on type of force, but not at a system level. And you can usually find some of that data from selected acquisition reports for major defense acquisition programs. But there aren’t any of those for ground-based radars. But the Missile Defense Agency does have operations and maintenance data for its TPY-2 radars. So to calculate the sustainment for the over-the-horizon radars and the PAD radars, what we did was calculated an acquisition to sustainment cost ratio for yearly sustainment that found that yearly sustainment is about 8 percent of the acquisition costs. And then used that methodology and applied it to the acquisition cost that we had for the over-the-horizon radars, as well as the PAD radars.

For the basis of the PAD radars, we used estimates for the Army Sentinel as sort of the cost baseline and the basis for the majority of the radars. And then at the center of the PAD, we had an additional more – sort of a little bit more of a boutique radar capability that’s an Aegis fire control radar. And that, at the center of the PAD, we got an additional estimate from a subject matter expert interview. For the aerostats portion, we had only one orbit of them deployed in the NCR, and used the more conservative of the CBO estimates, including their assumption that you would need to do sort of a reacquisition after 10 years. So when I flip over to the budget chart in a second you’ll see the sort of spike in acquisition, I think, in, like, years 16 or 17. And that is the reacquisition of aerostats.

For the Wedgetail radars, we didn’t have unit costs for the U.S. configuration because it’s still an RDT&E program. So we had to derive a unit cost from the U.K. acquisition program. So they bought five aircraft for approximately $2 billion. We assessed that the mission would require about two aircraft. And so that’s what the acquisition and sustainment denominator is there. And then for the sustainment costs, we used estimates from the P-8 Poseidon aircraft, which uses the same airframe as the Wedgetail, and then tacked on a little extra cost because it’s got a more capable radar on top of that.

And then finally, the layered shooters command and control and system integration components. I kind of rolled those all up into one line. For the medium-range interceptors that you saw on the earlier charts, we based our estimates for those costs on the Patriot M903 launchers and the PAC-3 MSE interceptor. For the long-range interceptor, we used the SM-2 Block IIIC. And the multi-mission long-range interceptor, we used SM-6 data as the basis for those calculations. Those were used to just kind of give us a hard data source and some sort of, like, formulation, rather than a specific endorsement of any of those particular capabilities.

For the VLS then, for the launcher at the center of the PAD, we also had to use data from ship-based VLS. The budget documents for, like, the Aegis ashore or any of the Army formulations that are launching – or, have a land-based VLS capability. They didn’t have specific unit cost data. So we may be purchasing a little bit of extra VLS capacity in that sense, but that’s the best we could from the public data. And then for the sustainment portion of that estimate, we used more – the more conservative CBO estimate for their LR-SAM sites of $20 million per year. The one element that we weren’t able to produce a cost estimate – and it’s included in some of the description of the Phase III – is some of the high-powered microwave capabilities that are discussed in the report. Since those systems remain prototypes, there’s not really an analogous system for me to be able to publicly provide a confident cost estimate at this time.

All right. The numbers are gone, although numbers did feed the chart. (Laughs.) Another important distinction between our cost estimates and CBOs is how we sort of think about the time horizon. So the CBO doesn’t really bound its time horizon. It has 20 year – just an overall 20 year sustainment cost, and then an acquisition cost. We tried to look at it from a bounded time horizon, and in particular because I think it’s a better measure of affordability to look at what the average yearly cost – or, be able to break it down by sort of a yearly cost, rather than the total cost. The total cost can produce a significant amount of sticker shock in many cases, but it’s important to know what the denominator for that number is. What are you dividing that number by? Because that determines, to some degree, what the early investment level is and whether or not that’s a sustainable investment.

So this chart depicts how the cost of the systems gets spread over that 20 year period. We tried to spread it evenly as possible. And that creates a little bit of a funkiness to the phasing, mostly because Phase II acquires so many assets. So Phase I is years one and two. Phase two is years three through six. So extending that out and doubling that period of time. And then Phase III acquisition is years seven and eight. And again, it’s sort of a rough estimate of the timing. It could probably be smoothed out even more as you’re working through the programmatics. But I think it’s important to sort of display and think about the costing in that way.

And it also shows why the sustainment number – the total sustainment number may look a little bit low, because sustainment get phased in. And especially because the phasing is a key part of our architecture. And so that means that Phase III assets have fewer years of sustainment counted in the time period than Phase I assets. The final architecture, we were able to estimate, once you get through all of the Phase III, has about a yearly sustainment cost of about $1.2 billion per year. So if you calculate that cost over 20 years, the same way that CBO did, it would be about $24.5 billion.

All right. And then it’s important to acknowledge that cost estimates are just that – they’re estimates. And they can give you a sense of the general level of investment, but no architecture is going to survive contact with environmental impact reviews or the first time that you have to have a system integrations test. It’s going to be incredibly difficult, and so it gives you a ballpark, in a sense. But it’s important to think through some of the contingencies and the predictable things that we can think about that are plausible alternatives.

So the first thing that is a plausible area and a potential area for cost growth is in the towered radars. We based it on the Sentinel radar, which has one spinning face to produce 360-degree coverage. If instead you need three faces or a multifaced design because, for whatever reason, the spinning configuration doesn’t work out, then that’s a potential area for cost growth. And then the other thing is we based our estimates – assuming that existing cellphone tower, phone tower, et cetera infrastructure would be able to support some of these radars, and so that reduces some of our MilCon costs, but if all of that has to be built, obviously there’s an additional MilCon cost, and that’s a potential area for cost growth.

Another area that we looked at in the appendix, and that the map image demonstrates there where costs could be potentially trimmed, where you have fewer overall OCHR locations, it will require a similar number of radar arrays. But if you can co-locate them in such a way that you trim up some of your sustainment costs and potentially some of your MilCon costs, it’s an opportunity for some cost savings. We were relatively aggressive and conservative in terms of choosing the PAC-3 MSE as our medium-range interceptor configuration, in terms of the cost. That’s probably a higher cost interceptor than you would potentially need.

And so if you look at other options like NASAMS or the Army’s future IFPC, those are other areas where you could potentially save some of those costs. And if you had fewer types of interceptors, in particular, you might be able to trim some sustainment costs in terms of the opportunities for finding economies of scale there. And then the final thing is directed energy. If it is able to mature substantially and it is able to reduce your requirements for the number of kinetic interceptors that you need, it can both reduce the cost per shot, but then also potentially, depending on the sustainment of those configurations, reduce the sustainment and the number of sites that are required there.

So with that, I’ll turn it over to Lee to begin the panel discussion. And thank you, again, Lee.

Ms. Hudson: Thanks for your remarks. I wanted to start off the discussion – the report points to warnings from Dennis Gormley about the emerging cruise missile threat. And I just wanted to dig a little bit deeper about why the United States is no longer a sanctuary, and why can’t current platforms like the North Warning System handle that threat?

Dr. Karako: Well, let me take the first part and hand it off. Dennis Gormley, you’re referencing a scholar who’s kind of been warning about the growing cruise missile threat for probably 20 years. He unfortunately just passed away, but, you know, I think it’s – we give him credit for noting this for a long time coming. I suppose the infinite charity that the afterlife brings probably is restraining him from saying I told you so, but there’s been a lot of people who have said I told you so that we’ve got to keep an eye out for this, and so – really over the past decade, and that’s kind of where we are today.

Col. Strohmeyer: And Lee, I would just add to that, the North Warning System definitely served its purpose for the threat that it was built for, essentially looking at bombers at 36,000 feet coming up over the Pole. What it wasn’t built for necessarily was looking for low radar cross-section targets that are flying at low altitude over the same area, so I think both Canada and the U.S. have both said, we’ve got to look at modernizing that system to allow us to be able to see the threat that has changed, based on what we see today.

Ms. Hudson: And another question I had about the architecture: You dig deep into the different elements, the technology elements, and I was hoping you’d talk more about what actually exists and what do we need to develop, what does DOD need to partner with industry to like actually create?

Col. Strohmeyer: That’s a great question. So I think General VanHerck recently said that he is looking for industry to take this problem and just run with it, creatively, though – and we would strongly agree with that. I think what we tried to build was not a pie-in-the-sky design or something that may exist 10 years from now, but we used almost all existing proven technologies and we just used them or we applied them in different ways, with new paradigms. So, for example, towered radars aren’t really that new but networking them together and using them in a different way can really provide that coverage that we might want but with systems that exist today, rather than waiting for. There are some elements, like high-power microwave, that are still kind of a little bit on the emerging side, and then certainly we wanted to build this out with the toothing to get that multi-mission application down the road for hypersonics, small UAS, and then also the ability to bring in emergent sensors, like space-based sensors that will be such an important part of the hypersonic tracking layer of the future.

Dr. Karako: I’ll just jump in a sec. We would have loved to have a bunch of directed-energy lasers on every pole kind of an architecture, and it would have been great to have said, oh, well, the space sensors will take care of that. But back to the Costco thing. You know, we’re looking at things that are on the shelf today. We can’t wait around – this is the temporal element. We can’t wait around until all those whizbang stuff, you know, becomes a retail item. And so we went – this is where I’ll be self-critical again. We went relatively conservative, less exotic. It would have been great to have those things, but we wanted to be the sort of thing that we could act on right now.

Ms. Hudson: And a question I had for Wes: When you look at the CBO report, the low end is 77 billion (dollars), and then you all come in way lower than that, and is that because you’re looking at existing systems and preferential defense?

Mr. Rumbaugh: I think that – yeah, I think that – I would say that the majority of it is the preferential defense part. I think that that’s the biggest cost driver is deciding how much you’re going to defend. The contiguous United States is very large; that’s a lot of square footage, which means that there’s a lot of assets that you’re having to deploy And so it runs those numbers up there very quickly. And then the other thing for a lot of the CBO numbers is aircraft; they’re expensive to operate and they’re expensive to build. And so – I mean, that’s something that we have to do in away games because we need mobility and we need the ability to operate. But in the U.S. homeland, we should sort of take advantage of the fact that it is U.S. territory and that the government can use sort of more ground-based assets to limit its costs more effectively.

Ms. Hudson: And, you know, you all mentioned in the news there’s so much talk about defending the United States against hypersonic weapons, but I was hoping that you could talk about what we’re learning from Ukraine about the cruise missile threat.

Dr. Karako: Well, I think, again, we are clearly thinking about the advanced hypersonic cruise missile or glider threat, and it’s why our element selection here was very clearly tied to the Mark 41 launcher for that Aegis-centric system. Why? So that when, in the fullness of time, the Glide Phase Interceptor or whatever else that is coming out of a Mark 41, arrives, that if the threat demands it that we’re able to build that in.

But in terms of the Ukraine side, you know, I think what we see there is – well, to quote John Plumb – that the weapons of choice are the slightly less exotic thing. Yeah, there’s some demonstrations going on, the whizbang Kinzhal, you know, quasi-ballistic hypersonic this, that, and the other thing. But there’s lots and lots of cruise missiles, and that’s what’s here today and I think that’s why you’re seeing General VanHerck, like, when he testifies, he kind of draws a bright red line under that Kh-101, right? That’s what Russia likes to publicize, that’s what they like to shoot off from the Caspian Sea into Syria or, according to news reports, they shot one into Ukraine the other day. And so these are garden-variety subsonic cruise missiles, and that’s where the threat is today.

Ms. Hudson: I also wanted to touch on – after reading the report – defense of Guam. It seems like the Pentagon is all in in defending Guam. What can we learn from that and apply it to homeland cruise missile defense?

Dr. Karako: So I would say that while we did not build – deliberately did not build a nice, pretty picture of Guam, we could have put towers and such on Guam, but we didn’t, deliberately; we want to kind of see how that ripens. But we’ve already heard from a number of DOD officials and from, for instance, the Missile Defense Agency, that – you know, that’s a hard problem and it’s going to have applicability here.

Couple comments on defense of Guam. Based on news reports and public statements on Capitol Hill and, frankly, here, it sure looks like it’s going to be a mix of Army and Navy assets. And I think if you look at our architecture here, you see a fairly straight line between the mix of assets that are being publicly discussed for Guam and the mix of assets and the sources of the assets for this architecture. You also see, of course, that we add in that extra money for systems integration. Assuming that it’s not all going to be worked out for Guam, we want some additional funds there to do that. So that is – Guam is, of course, U.S. territory; it’s the beginning of the homeland cruise missile defense undertaking. But I would also, again, point to the thing being protected and the strategic logic of the defense of Guam which is it’s not just about the beaches and the people, although of course that’s important, but it’s the ability to project power from those airbases and other things on Guam. And so that’s also a lesson here is, what is it that we’re trying to defend and for what purpose? And that’s why, again, you see the head of TRANSCOM and the head of NORTHCOM and STRATCOM highlighting those kinds of things.

Ms. Hudson: And say that DOD and lawmakers want to invest in something like this in fiscal year 2024 like you outline in the report. How soon could this become reality? Like, what are you expecting?

Dr. Karako: I’ll let Wes take some of that, but I’ll just say real quick at the beginning that I certainly hope that PB24 has a very substantial piece here. It’s not just about cruise missiles again. It’s about the broad air threat, it’s about the broader surveillance and detection and attribution and everything else, so it’s not just about that one threat.

But let me turn to Wes in terms of the procurement timelines.

Mr. Rumbaugh: I mean, I think that that timeline, sort of going back to the answer that Tom just gave, kind of depends on how much the architecture mirrors that of what’s going on on Guam and how much you’re sort of getting out of those non-recurring investments in architecture on Guam. But, I mean, that was sort of the point of not looking at the whizbang investments in relying on things like directed energy is, you know, you’re talking about surface-to-air missiles, ground-based radars. These are the sorts of things that U.S. industry and the United States is already procuring and so it’s changing the number on the order sheet for capabilities that are already existing today, rather than attempting to invent a bunch of new things.

Ms. Hudson: What, in terms of initial investment, do you – should, like, the PADs be funded first? Like, what are you looking at?

Mr. Rumbaugh: I mean, I would kind of defer to Matt in terms of – like, in terms of the sequencing, we kind of sequenced it in terms of the phases for a reason, and so they’re sort of – I think that the nature of the architecture – and Matt, you can correct me if I’m wrong – is sort of such that all of the investments are necessary together because that is the point of the layering in the defense in depth is to invest in all of the capabilities together at the same time, and so I think that you would probably need to invest in most of the capabilities. Yeah.

Matt, do you want to jump in?

Col. Strohmeyer: I was just going to say I think certainly we heard General Roper say it and I think General VanHerck has said it many times that we have some serious domain-awareness challenges right now and shoring those up for these emerging threats that we see today are important and that’s why, like Wes said, we phased it in – yes, the five layers we think are all important, but we phased it in a way that gets us capability very soon in the next few years for some of those, like, domain-awareness for the northern approaches, for example, OTHR, and then also, though, putting down a real, credible detection and defense capability over, like, that first PAD, so that you have real deterrence that’s actually being put into the field. And we think that’s not only important for deterrence, it’s also important to show, hey, this is doable, you can do this, we can get the capabilities in the field and learn from it, and then, over the course of those phases, expand on what we’ve learned from and expand on that defense deterrence capability.

Ms. Hudson: One of the interesting ideas I thought in the report was partnering with the FAA and NOAA and other organizations to get that domain awareness. Is that something that DOD has done before? Like, what kind of steps would they have to take to get there?

Col. Strohmeyer: So I’d say that – so DOD partners very broadly with the interagency. Every day, like, NORAD defenders, whether they’re Canadian or Americans, sit almost side by side within the interagency as their main defensive awareness for North America, but I think there’s potentially opportunities to look at how we think about whether it’s sensor data, for example. We often think about it kind of stove piped into the mission that it was designed for – for good reason; it’s designed for that mission – but trying to look at opportunities to combine some of that data together in ways that we might not have been able to do in the past so that, for example, if you have a sensor that was built for a mission, maybe a counternarcotics mission or a border patrol mission, if you can take that same sensor and combine it with other sensors, even though that’s only a small portion, you can start building out an aggregate that turns out that you can see potentially more than we thought we could and for relatively small investment, not building new sensors but taking what currently exists for that layer three, that wide-area surveillance there.

Dr. Karako: And that’s what makes this, I would say, very near term, a unique opportunity and test case for the department’s concept of integrated deterrence.

Ms. Hudson: I was going to go to some of the audience questions. One question is, what will be the role of DOD’s Missile Defense Agency and the concept of homeland defense against cruise missiles?

Dr. Karako: I’m just going to point to the HASC language about – for the forthcoming NDAA, reminding – gently reminding the department of the FY2017 requirement to identify an executive agent. That decision has not been made. I think it will probably be made soon, based on some comments made on Capitol Hill, and that will ascertain, you know, who’s going to be doing what in the scheme of things.

Ms. Hudson: OK, so we have to wait for that. (Laughs.)

Another question: Given the surprising concern of the U.S. being endangered, are we looking at the rise in cruise missiles? And also, are hypersonics going to be the certain future for such weapons?

Dr. Karako: You know, I’ll recur here to some themes that we put out in our – since the audience member talked about the hypersonic stuff. When we put out a report in February, called Complex Air Defense, we called it that specifically to accentuate the aerodynamic endoatmospheric quality of the threat, and we said, look, put aside the whole whizbang hypersonic stuff; just take note of the fact that the threat is going lower and is becoming more maneuverable. And that’s kind of what ties all this together, and it’s also – again, back to Deputy Secretary Hicks’ injunction to don’t just think about one thing, think about the multi-mission thing – that’s why you see the element selection here, that’s why you see the toothing for counter-UAS and hypersonic in our architecture today. That has to be the right solution. We can’t develop a stovepipe for this, that, and the other thing, and we have to be smart about beginning that from the outset.

Ms. Hudson: I think that’s about all the time we have. Did you all have any closing comments?

Dr. Karako: I would just say stay tuned and we’re going to have a great couple panels here from government and from industry to tell us kind of where things stand and where things could be going.

Ms. Hudson: All right, thank you. (Applause.)