CDE Brief | Panel Discussion: Collateral Damage Estimation

Welcome to the Naval War College. This is the International Law Department's presentation on collateral damage estimation methodology. I'm Major Jeff Thurner with the U.S. Army. This is Major Tim Kelly with the U.S. Marine Corps. Both of us are faculty members here in the International Law Department, and today we're going to give you a presentation that's drawn from a class that we attended recently in the Joint Targeting School. From the onset, I want to make it clear that both of us are military attorneys. We are not experts in the collateral damage estimate or CDE process, but our goal today is to provide you with some basics so that you can kind of understand the general underlying principles and so that you can be conversant in the subject and understand ultimately what the process outputs are. So when you hear the final CDE call is CD level whatever, you understand what that means. Additionally, this presentation should give you an appreciation for the process and the careful approach that the U.S. forces take to eliminate collateral damage. Now other nations have developed similar systems, but today we are going to be focused solely on the U.S. system. As military lawyers who work in the area of law of armed conflict, we frequently discuss issues such as proportionality, distinction, and precautions in attack. The CDE process takes account of these issues and provides a structure for determining the appropriate level commander who will make the decision regarding whether, and in some instances, the method and weapon used to strike a target. We'll mention that throughout the presentation that some aspects of the CDE methodology are classified. This is an unclassified presentation, so we will not be discussing any of those classified items. With that said, let's take a look at the general CDE methodology. First thing we're going to start is show you the references. The primary reference is the Chairman of the Joint Chief of Staff's Instruction 3160.01. That really is the bulk. It shows you how to walk through this process that we're going to do today. There are some tables that are associated with that. Those are classified tables. We're not going to be referring to those, but we'll talk generally about how that process works. The reason that this methodology was developed in the first place is because there were some perceived problems in the 1990s when the U.S. was trying to conduct targeting. The issues were that estimates of collateral damage varied depending on what unit was doing the calculation, what service, what the location was, and so it became apparent that we needed to have a unified approach to this. As you can see, the quote from General Zinni, one of the major proponents of the development of this system, was that he wanted something simple, logical, and repeatable. Something that he could even do himself. So that if you're sitting in Afghanistan, or in Tampa, or in Washington, D.C., if you're given a certain fact scenario about a proposed target, you would come up with essentially the same estimate for the amount of collateral damage that would be caused. That was the genesis for it. When you take a look at the methodology itself, you'll see that it is now a logical framework, and it gives you a processor estimating these things. It uses empirical data, so there's certainly a lot of scientific research that's been done to determine this, to try to give you the best estimate for the amount of collateral damage that may be caused by a particular strike. This system is across the board. It applies to all services, and it also applies at all levels of combat. So from the tactical level up to the strategic level. The other key thing to remember is that this is for deliberately planned targets. So it's not for hasty operations or something where it's going to affect the commander's sensitive targets that he's interested in engaging. When you look broadly at what this process is, you look at the legal aspect of what this is trying to cover, it's not really a proportionality question that you're trying to answer. It's really more of a precautions and attack approach. It's showing the feasible steps that the U.S. is taking in order to minimize civilian collateral damage. It's also primarily about the level of command that can approve a various strike. So depending on how far down the CD process we get, depending on how grave the concern may be about collateral damage, it'll tell you what level commander has to approve a strike. Whether there's really no concern, it could be a lower level, but as you get higher up, it's going to require up to some of the highest levels in the national command ability to approve those strikes. So that's really a broad framework for what we're looking at. The reason we have this approach is because there's significant strategic risk if we fail to properly account for collateral damage and properly take steps to minimize the amount of damage that we cause from a certain strike. So to avoid those strategic risks, we have developed or the U.S. has developed this methodology that we're going to walk you through today. Before we get into the methodology itself, I want to cover a few terms and issues related to collateral damage estimation. The first being is what is collateral damage itself? 3160.01 addresses collateral damage as you can see on the slide as follows. It is the unintentional or incidental injury or damage to persons or objects that would not be lawful military targets in the circumstances ruling at the time. Such damage is not unlawful so long as it is not excessive in light of the overall military advantage anticipated from the impact. Any questions or comments on that definition at all? Is that definition the accepted definition internationally or is that specific to the U.S.? Well it's particularly derived from, like I said, from Chairman Joint Chiefs of Staff Instruction 3160.01 which is a U.S. document. So that is the U.S. view. I can tell you that from the point of view of additional protocol, one, that they address collateral damage in the following manner. This is a prohibition against indiscriminate attacks. An attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects or a combination thereof which would be excessive in relation to the concrete and direct military advantage anticipated. So you see with AP1, which most of the rest of the world would adhere to in the U.S. view, not a lot of difference, but you will see while both of them mention the direct military advantage anticipated, AP1 also includes the term expected to cause. Now that's, while not specifically addressed in the U.S. view, it's clear in U.S. policy that collateral damage estimation is calculated with the knowledge that the commander had at the time of approving the target, you know, carried through until execution on the target. This isn't something that's done after the fact with a battle damage assessment and then someone second guesses after the fact whether the actual damage was excessive in light of the military advantage gained. Sorry, I have one more question. Is there a difference between the overall military advantage and what I think you read from AP1 saying the direct, or sorry, the concrete and direct military advantage? Is that another difference? I don't know that it would be a substitute difference other than, you know, the use of a couple of different terms, but I think the... Well, I think it's a little broader though. It basically gives us a broader view. So you're able to look at the full battlefield and determine if this attack in relation to the whole battlefield is, if that attack would have a substantial, or would be, the gain is excessive, or the damage is excessive to that gain versus looking at that particular aspect of the fight. Because, for instance, feints or other things where maybe that the direct advantage of that attack is not so high, but in the light of an overall battle, well, yeah, the overall gain is significant. So we're not looking at just this fight against the damage caused. We're looking at how it fits into the overall context. So I think that that little nuanced difference is apparent in the definitions. Now, specifically, why do we conduct a CDE analysis and what questions does the process seek to answer? Generally speaking, there's five simple questions that need to be answered before engaging any target. First is, can you positively identify the target as a legitimate military target authorized under the current rules of engagement? Second, are there collateral objects or persons or significant environmental concerns within the effects range of the weapons that you'd like to use? Third, can you mitigate damage to those collateral concerns by attacking the target with a different weapon or method of engagement and yet still accomplish your mission? Moving on, fourth, if you cannot mitigate those effects, how many non-combatants do you think will be killed or injured in the attack? And finally, are the collateral effects of your attack excessive in relation to the expected military advantage gained? And do you need permission from a higher commander in order to execute that target? The geographic commander responsible for an area of operation promulgates two different target lists, the first being a no-strike list. And on this no-strike list, the all-identified objects within a specific geographic range which are functionally characterized as non-combatant or civilian in nature are listed on there. And as the name of the list indicates, these are not to be struck. Also, there's a restricted target list, and that's where a geographic combatant commander has imposed certain weapons effects or certain restrictions on different targets. And that could be for political reasons or intelligence gain loss reasons or certain policy reasons or whatnot. Now, we're talking about pre-planned targeting here. So if, for instance, someone wanted to strike a target that was on one of these lists, you'd have to go back to the geographic combatant commander. And if it was on the no-strike list, have it removed from that list by the competent authority and potentially moved to the restricted list or get permission to use the particular weapon you're looking at for the restricted target list. Another concept I just want to go over prior to getting into the methodology itself is dual-use targets. Dual-use targets are those targets that are characterized as having both military and civilian purpose. And the U.S. policy is they require a casualty estimation where you count everyone within that structure as a non-combatant civilian casualty. Why would you count everyone as a non-combatant if it's dual-use? Well, the U.S. directives require that, and really it just shows the extensive caution that the U.S. system takes in taking into account the number of collateral casualties you'll have caused by a strike. And we have to remember, we go back to this, this methodology doesn't necessarily approve – doesn't approve or disapprove a strike. It assigns the appropriate level of commander who can approve a strike. So you could run into dangers if you counted some of the people and not some of the people where you potentially have a lower-level commander approving a strike, where really because of the dual nature of that structure, it should be handled at a higher-level command. With regard to human shields as they come up, human shields are non-combatants that are intentionally placed near a valid military target to hinder the attack. You can have two different types of those. We could have voluntary ones – or excuse me, involuntary human shields, and those are people that are against their will, have been placed near a military target, and those must be accounted for in any casualty estimation. You also may run into a situation of voluntary human shields, people who potentially support one side or another, not necessarily involved in armed conflict itself, but willingly place themselves in harm's way in front of a military target in order to hinder the attack. The U.S. view is that while involuntary human shields must be accounted for in a casualty estimation, voluntary human shields would not be accounted for. Sorry, is there consensus in the international community about that point of view? I wouldn't say there's a consensus in the international community, but it is not uncommon that the other view would be held with regard to voluntary human shields, and that some nations that the U.S. operates with may count not only involuntary human shields, but also voluntary human shields in their casualty estimation. It's important to note, if the U.S. doesn't know the nature of the human shield, we treat them all as involuntary. So if our intel is not sufficient, that we know whether they're involuntary or voluntary, we treat them as involuntary. And then finally, before we move into the methodology itself, I want to just go over a couple things with regard to weapons effects considerations. First being, what are the primary causes of collateral damage? The first and most significant cause of collateral damage is PID failure. What that means is that although we believed the target to be of a certain functionality, of a certain use, of a certain nature, in effect there was some sort of error that the target was identified as a lawful military targeting error. And that's the majority of collateral damage caused. The second would be a weapons malfunction. Quite simply, the weapon doesn't work as it was designed or as it was planned. And then third would be a conscious decision. Now realize, when we say conscious decision here, we're not talking about somebody that's intentionally targeting civilians, because that wouldn't be collateral damage. That would be a war crime. The conscious decision we mean here, and you'll see as we go through the collateral damage estimation process, is that at some point we can expect that our weapons will have adverse collateral consequences, collateral casualties. And at the end of the proportionality analysis, the appropriate level commander makes the decision that the expected military advantage gain, that the collateral damage is not excessive in light of the expected military advantage gain and improves the target. So we prosecute that target with the knowledge that there may be collateral damage. There's also issues with delivery accuracy. The U.S. probably has the most accurate weapons in the world and continually strives with research and development to get more and more accurate weapons. However, we still have some issues. Target location error. We do still rely on humans at some point in pinpointing a target and determining exactly where it is. There could be a map error, what have you, that could add up to an error in the actual location of the target in real life. And then finally, guidance or delivery system variation. As I mentioned, the U.S. has made strides, the world has made strides in increasing the accuracy of weapons systems. However, they're still prone to some variation. Then finally, the collateral damage methodology does not account for weapons malfunctions. Quite simply, there would be no process if you tried to account for every different malfunction that could happen to a weapon. And it also does not account for unknown transient noncombatant personnel and their property in the area of a target. So does that mean, for example, if you knew, if you were somewhere in the middle of nowhere and you had intel that there were people coming through, how would that be dealt with? Right. Within the methodology, unknown transient noncombatant personnel are not accounted for. I think as we walk you through the methodology, why that is, because it is a mathematical, to some degree, system. However, if there were a known transient that was not expected to be in the area, not generally the usual person that would be in that area, but was known, that would clearly be briefed to the commander and incorporated into the targeting analysis, the proportionality analysis, the precautions and attack, all those sorts of things. Known intel is going to be accounted for and all that. This raises an issue, though, of time lag at this methodology where you're taking into account all these factors. And it comes to mind when you talked about these unknowns that might enter into the sphere that you're operating in, in terms of the target zone, I guess, or the targeted object, and the fact that is there a time delay there that you go through that methodology, there is now this room for the environment to change, so the transient is going to enter. I guess maybe you'll get to this in terms of the time factor, that this methodology takes time, and then, you know, does that time affect what then happens in terms of the target itself, or is that accounted for in that? Well, ma'am, I think what you'll see is that the CDE takes the target and looks at it as it would normally be and accounts for different variations, night, day, and we'll talk about the episodic events in that. But it doesn't necessarily seek to provide an output that reflects real time. It reflects what can be expected under the general circumstances that prevail in the area of that target. So if you had particular intel, it would opt you out of that, and instead of looking at something as you expect it to be under normal circumstances, it would be analyzed under what we know it to be. For instance, if we had some sort of surveillance on the target or something like that, that would potentially, as you think you'll see as we go through the analysis, somebody would almost opt you out of that type of CDE analysis as opposed to, or put you into one where we know the actors in the area. Right. I mean, I think it's continuing duty. So if you now know of somebody in the area, then that would change the analysis, and then we don't continue to go forward on the strike saying, well, beforehand we thought there was nobody, even though we know there's somebody now, we're still going to engage because we already had approved at this level. So it would now change the character, so we'd have to go back and get new, if you had a new person in the area. The idea about not taking into account the unknown transients is that we don't have a fudge factor. We don't say, oh, maybe somebody else is going to happen to walk through this area, so we'll plan for that contingency. We don't plan for those contingencies. We plan for what's known. And so if we don't know of anybody, then we'll make an assessment based on that. If intel later on comes and shows that that was incorrect, well then, yeah, we'd absolutely take that new intelligence to account and have to go re-look at our target engagement at that point. Maybe you'll talk about this as well, but how long approximately is this process? Sure. I mean, it varies depending on your area, but for instance, operations that are ongoing in Afghanistan, they have teams that have CD analysts that have been pushed forward into theater from Central Command who are making those calculations. And you're they can get those counts down to 15 minutes. I mean, it can be a relatively short period of time depending on the situation, obviously, in order to make that call. Okay. Well, now we want to take a look at the methodology itself. And we're going to take a look at the various levels of what CDs and before I even talk about the levels that I guess I want to talk about the process and that there's two different processes that are happening. You have a weaponry process where a target is identified and then the weaponeer is going to look and see what weapons are available, see what weapons make most sense for this circumstance. And it's going to make a recommendation based upon that, that they want to attack this target with this particular weapon. So they'll identify the precision-guided munition, the aircraft that they want to use, all those sorts of things. And at that point, when the weaponeering is done, it will then be passed over to a CD analyst who will take a look at the situation and make sure and identify what the CD level is appropriate for that call. And then you'll be able to see which level of command then must approve that strike and whether in fact the recommended munition that the weaponeers come up with would in fact give you, you know, would give you a level that you'd want that you'd want to use that or there may be recommendations to use a various other type of weapon. The CD analyst will help that part of the process. There's kind of two things happening subsequent, I guess, but somewhat also a little bit in parallel. They can kind of go back and forth to help get you to the point that you're looking to so that you have a definite understanding of who can authorize the attack and what is the best method for doing that. Okay, when you look at the CD portion, so now after the weaponeer has given us that information, the CD analyst will take a look through five different levels and he will look to stop obviously at whichever level he is confident that the collateral damage concerns have been identified or have been mitigated enough that he can then approve it. So we're going to, I think it'll be clear once we kind of walk through. There's two things that happen when you first begin the process. When you first get the information, you're going to, the CD analyst is going to make sure that in fact the target has been identified and that there is a geospatial identification of the target, so you know what you're looking at, what house you're looking at, what building or structure. And also that there is enough information for that to be positively identified as a valid military target. So you're looking at those kind of background issues. If you don't have that, if for whatever reason you aren't clear which house it is or you're not clear if the target is in fact a valid military target, you would stop at that point and you would not go forward. You would not be engaged. The CD analyst would stop it and say we're not going to engage that target. Similarly, the other thing they're checking for is whether there are any law or violations in attacking that target, whether there's any ROE violations. If you have those sort of things, then again the CD analyst will say no, we're not going to attack this because we're not complying with those rules. So that's kind of an initial check. The other thing they're checking early on is whether there is any sort of dual use. If this is a dual use facility that may be damaged or will be attacked, whether there's a chemical hazard, environmental hazard, any of those sort of things. If you have those, then you know that ultimately you're going to get to CD level 5. Any of those three things is going to lead you, no matter where you are in the process, if you're able to stop earlier, basically you aren't because you're going to have to go to CD level 5. That's kind of mandated through the system that if you run into those. Okay, with that basic understanding, let's walk through the methodology itself and hopefully become a little clearer as we talk through it. Now first we're going to look at CD level 1. This is the initial target validation, the initial assessment. The CD analyst will be given the target. He will take a look at, do those first checks we just talked about, and then he'll take a look at where that target is located on the ground and what collateral concerns are in the area. He'll take a look and he will establish a collateral effects radius. Now the distance for that radius, that's part of the, you'd use the classified table. So there's a classified number, how many meters away from the target you'd be looking for, and this is for all attacks. CD level 1 is the same for whether you're using precision guided munitions or unguided munitions or surface-to-surface munitions. There's a fixed amount. Clearly that's a pretty wide area, so we're looking at many hundreds of meters out. He would do a box around the target and ensure that nothing falls within that. And in this case, as you see here on the first slide, you'll see that the collateral concern number one falls outside of that box. So when the CD analyst is looking there, he sees there are no collateral concerns in the area, and so he's able to make a CD 1 low, which means yes, you can approve the strike at a CD level 1, which means whichever commander is authorized to approve CD level 1 can in fact authorize that attack because there are no collateral concerns that are within that collateral effects radius. And that number, the distance that is around that building is essentially what science has shown. The biggest of the bombs or the largest risk that we have, that's the value that we use for CD level 1. So take the worst case scenario essentially. What is the worst case scenario? As long as nothing is within that worst case scenario box, well then we'd be authorized at CD level 1. As you see as you build, if there was in fact a collateral concern that was inside of that box, now we would have a problem. We would not be able to approve that strike at CD level 1. So now we would be forced to advance to CD level 2. And at CD level 2, we're going to start to look at some of the other aspects of the round. We're going to look at the class of weapon. So when you take a look at CD level 2, there's three different approaches. This is really where it separates into three different approaches. We will either follow the approach of precision guided munitions where we follow the 2A, 3A, 4A path. Or you look at a path of an unguided, an air-to-surface unguided, or a surface-to-surface ballistic munition. Now again, I told you the Weaponier is already giving you this information. So the Weaponier has specified in advance of the CBA as getting it, we intend to use this particular weapon. And then you would follow those particular paths. Now today we only have a short amount of time. So we are going to focus solely on the precision guided munitions. We are going to stay on that top track. The other two tracks, the air-to-surface unguided and the surface-to-surface, those tracks are somewhat similar. They have some differences. Certainly they use different values for the classified tables applying to those. The one thing I will tell you about those systems is in the CD level 2, the first check really is a feasibility check. And if it's determined it's not feasible to use those types of munitions, there's a recommendation back to the commander at that point that the commander should instead look to use precision guided munitions instead of either the unguided or the surface-to-surface. So that's the first step. But anyway, we are not going to discuss the rest of those processes. We are going to stick solely to precision guided. So let's take a look at the example, as an example in CD level 2, where we are looking at a precision guided munition. So the Weaponier has told us we're using precision guided munition. Now we take a look. We have our outer box, okay, that CD level 1 box that showed us the distances that were appropriate. Now what we do is we take what is the worst case scenario for a precision guided munition. What would be the biggest concern for any precision guided munition? We're not looking at a specific type of weapon here. We're just looking at the class of weapons. So we're looking at all precision guided. So you take the worst case scenario for a precision guided munition and you essentially draw a radius around the target and you look for collateral concerns. Now in this particular case, you see that the collateral concern is outside of that CD ring. So you're able to make an appropriate call, the CD analyst can make a call there, that this target is authorized under CD level 2 low, as long again as you're using the precision guided munition, the PGM, and in this case it's a unitary strike versus a cluster strike. So a unitary strike on that target would be appropriate at a CD level 2 call. But if a collateral concern, as you see with the next build, if the collateral concern is actually within that circle and even though it's just barely in the circle as you see here, if any part of that lies within that effects radius, then we're not able to make a CD level 2 call. You're now forced to continue on to CD level 3. So it just takes a small part of that building to be within that circle and it's going to bump you up to the next level. Okay, when we take a look at CD level 3, now we are focusing on the specific type of munition that's used. Again, the weapon here will have provided that information to the analyst, told him what particular munitions he's going to use. Again, we're staying all within the realm of precision guided, so all the things we're going to talk about today are precision guided munitions. So in this particular case, the weapon here has told the CD analyst that they would intend to use a Mark 82 round, a particular type of a precision guided munition. So we now analyze this under CD level 3 and there's two different things that we look at in CD level 3. You look at an unmitigated and a mitigated, but are both specific for that type of munition. So you look at how damaging is a Mark 82 round is essentially the concern. In science, the tables, the classified tables that are provided as part of the provided for the CD analysts will tell them based on that research, how much damage can be expected to be caused from a Mark 82 round that lands on that target. Unmitigated means it's just dropped as is and so we would have a circle as you can see on that slide that shows an unmitigated ring. So it shows what sort of damage would be caused regardless of if you just dropped a Mark 82 onto that particular target and you'd look for collateral concerns. In this case, there are none within the circle, so you would be able to appropriately make a CD level 3 low call with a precision guided munition Mark 82. There would be no issue. The concern would be what if that collateral concern was actually within that unmitigated ring. Now you look at ways to mitigate the effects of that round and you can mitigate through doing things like air bursting, shielding, offsetting where your aim point is, adjusting the heading and one of the main ones that's used is delayed fuse. With a delayed fuse, what you do is you ensure that the round when it hits that target, it will not explode until it is inside of the structure of that target. So it will cause an internal explosion in the target building which would minimize the amount of collateral damage that you have or minimize the amount of damage that will occur outside. So you end up with a smaller ring instead of that big outside unmitigated ring. You would deal with a much smaller mitigated ring and again the tables would tell you what that amount is. The science will tell you a delayed fuse which goes inside of that structure and then explodes or if it missed the structure, a delayed fuse would also ensure that it's buried in the ground before it explodes. So you're going to have far less of a concern. So you'd have a much smaller rings you see here on this next bill. Now you look at that mitigated ring and if in fact as this case shows that there is no collateral concern within that mitigated collateral effects radius then you'd be able to make the call of a cd level 3 low pgm with a mark 82 and then you put the caveat on there that it's only with the delayed fuse because if you weren't using delayed fuse as you see it would be within that unmitigated ring. But so here we would have no issue. Here the next bill showed you that unfortunately here the collateral concern would be inside of that mitigated ring. So even when you've taken all the steps to delay the fuse to try and shrink that circle as small as you can you still are going to have a collateral concern that is affected by this attack. So now we force at that point to proceed to cd level 4 which gives you a more refined assessment. You take a look at cd 4 a couple things that we're looking at in particular now we are mainly we're primarily concerned the main focus is a indoor it's an indoor concern. We're worrying about the blast damage to the surrounding buildings the collateral buildings outside and because of that we're focusing on what is the composition what is the makeup of that particular building that's closest to you. So you're worried about is it a wood building or is it a brick building because those buildings are affected differently based on the blast. So I'll you how that works. So as we go to cd level 4 the analyst is going to take a look and he needs to know what sort of structure that is what is the composition of that building. So in this particular case the cd analysis is informed that it's a multi-story masonry building. So it's a brick building it's a couple stories tall so you can imagine that it would be a rather you can expect it to be much more substantial to be able to withstand some sort of types of certain amounts of blast damage without being negatively affected and all of that has been calculated and that's what the cd level 4 tables would tell you the classified tables will tell you how how much damage would a mark 82 that hits that target how much would it cause for a multi-story masonry building that those values would be identified and so it would give you a particular radius. So you'd look at it and if in this case you had a multi-story masonry building at that distance away from a target again all the distances themselves are classified but you can assume in this case as a picture as picture shows that it's outside of the effects radius of a mark 82. So you would end up with a cd level four low pgm strike again again with a mark 82 delayed fuse which you're doing to mitigate damage and then you're also adding in cd level you have the heading so the heading goes from zero all the way around 90 180 270 360 so it's showing you that you want the aircraft to take off and go off to the right hand side of the target because it has often uh or it is sometimes there's a tendency for some of the rounds to go long if the round were to go long it would avoid all the other collateral concerns that you see on the left hand and the bottom side of the of the target itself so you have the head so in this case you would have no issue uh the cd annals would make it a cd level four and with those caveats and the commander at cd level four could approve the target could approve a strike on this target based upon that but if instead of it being a multi-story masonry building if instead that building is just a single story adobe building so it's the same distance away and the same distance away but it's a different composition of the building so instead of being a nice sturdy brick building you have just a single story adobe building well that building is going to have is not going to be able to withstand the impact the same as a brick building would and so the cd level four tables would give you a very different value if you knew that that building was a adobe building and it would in fact give you a circle as you can see here which is far outside of that that circle that we had for a brick building so in this case now that it's that you're again you learn that it's a single story adobe building you would not be able to approve that strike at cd level four it would have to go on to the cd level five and so you would need a higher level commander to approve that strike again cd4 is that really affects uh what's where damage is caused by the strikes and then would lead you to believe in this case that you need to go to cd level five which is the last level and we'll talk about cd5 when you get to cd level five you are assuming now that you are going to cause collateral damage that there will be uh unfortunately people killed so as much as you've mitigated it as much as the other steps you've taken to try and make it tight as circle as possible the surrounding buildings the collateral buildings based because of their type and because of their proximity to the target structure they will be damaged and you can anticipate that people would be killed so now we're at cd level five what we do is we look at the surrounding buildings and we look at what they are used for so last in cd4 you're worried about the composition now you're worrying about what is the building what's the is it a house or is it a market is it a warehouse you know those sort of things because under cd5 we the u.s has developed population density tables where they tell you based upon what the structure is used how many people can be expected to be in that structure during the daytime or at night or during some sort of episodic event and so you'll have a differing number of people that you can expect to be in that building based upon what it's used for and then you compare ultimately we'll walk you through how you do that but ultimately when you're given a number of expected casualties you then compare that against a limit that the commander has established for being appropriate in these circumstances now one thing to understand about the cd process and particularly about the casualty estimate is that it's not an exact science there you know it uses science it uses a lot of research to try and develop this but it's impossible to it is the best estimate though that the u.s has been able to develop so let's take a look at an example of cd level five so we have a target and we have two different collateral concerns that are both within that that last circle that we looked at the cd level four so we were unable to mitigate away from them we know that we're going to cause damage to the adobe structure just to the north of the target and also to the masonry structure to the west but now again as i told you level five we're worrying about what are the facilities used for okay not just what the composition is but in five you're worrying about what is the purpose of them so it's important to know that the one to the west is a warehouse and that'll affect how many people you can expect to be there on the daytime at night and on any sort of episodic event and the one to the north is a single family urban house so based upon that it'll help you determine what what you can expect for how many people you can expect to be inside of those homes at those various times okay the other thing that we look at is that now we look once more at the proximity to the target and you basically take that smallest circle that you had already developed from cd4 and you cut it in half and that the inner circle is called the inner annulus and the outer ring then is called the outer annulus the assumption is that anything that is in the inner annulus so it's that close to the target you can assume that when the strike hits the target that everything within that smaller circle essentially would be killed and so you would be using the casualty factor of one you'd be assuming that everything in that structure unfortunately would be killed but in the outer annulus because of the way blasts dissipate over time or over distances you can expect a smaller portion of people in those outer rings to be killed so you're coming up with a in this case 0.25 will use as being the factor of the type of the amount of people that would be killed out in the outer ring are you looking also at what potential damage might happen to objects as well as cde the into account the the estimates are all based upon casualty so they're focusing on the number of people that are that are killed in the strikes so the cd5 is a strict casualty estimation so it's going to tell you the number of people that would be killed and essentially the process is looking specifically at people that are going to be killed and certainly things such as buildings that are going to be damaged will be factored into it but it's mainly looked at to know if we are going to kill innocent people and that's what we're concerned about that that's what kind of drives the level of approval it needs to the level of the commander that would have to approve a certain strike it's based upon the number of people potentially okay so you have which helps tell you it helps you it's going to help you define or determine the amount of unfortunate people that would that would be killed in the strike the other thing that you're worrying about is the size of the structures and the amount of the structure that's actually um the amount of the structure that's affected by a particular attack so you you would have imagery handles that would tell you um how much of the building is actually covered by the strike so if you see the the structure to the west there you see that only a percentage say 40 percent of that structure is affected when the strike hits the target that portion of the building would be affected and so we'll use that percentage in our calculations because not everyone in the building would be affected only the people in that portion whereas the building to the north cc1 that you see on the on the on the slide you would anticipate that everything in that building would be affected so we're going to use the entire 100 of that building okay so essentially when it's established we have a little table that we've done a little worksheet for you at the top and it shows you the breakdown so we have the two structures cc1 and cc2 first is the single family urban home and the second is the warehouse we look at the total square footage of the facility and how much of it is affected with the with the house all of it is affected so we're using a factor of one so all 700 square feet are affected then we look at the population density tables and you're looking at how many people in this region and one thing I did mention but it is important I'll mention it now is that for the population density tables they are done very specifically per area of operation done per country these are things are established so it's in that particular area how many people live in a home how many people can be expected to live in a single family home and it would certainly vary depending on what part of the world you're in so again we look at our particular area that we're focusing on this target how many people are expected to be in that single family home in this case based on the for instance we'd say that one person per thousand square feet is in that building during the daytime where three people per thousand square feet would be there at night using those calculations we'd end up with if we did cause that we did attack that target we would end up having casualties of one person during the day and three persons at night if we attacked the single family home then you look at the warehouse cc2 you look at the warehouse it's a much larger structure it's but only 40 percent of that building is affected so the affected area is 2,000 square feet and you look at the population density tables which tell you generally per thousand square feet you have one person in the in the warehouse per day and half and essentially half at night and since it's 2,000 square feet you end up with totals of two people during the day and one person who would presumably be expected to be in that building at night time so if you attacked you as you see on the worksheet you'd end up with totals of three people who would be killed in this strike between the two buildings if we attacked a day you would have four people who would be attacked at night that's your cd5 casualty estimation based upon that you would then look at what the commander has authorized the strike what what is the non-combatant value for this strike and if the casualty estimate you take a look at it if that number you compare the two if the strike if for instance the commander had authorized attacks that were that would cause four casualties or or fewer then you would be able to attack at day or at night but if the commander only authorized attacks that cause three or then under a five series you would only be able to do that one at a five low anytime it exceeds that amount it goes to a five high and that's that's essentially where we get a five and again it's again a differing commander based upon a five low or a five high who has to approve the strike and we're going to walk you through one full example tim will walk you through an example where you start from start to finish so you can see how this process works how you would go from one up to potentially five depending on the casualties okay so let's take a look at a proposed target and a proposed weapon that we want to use use against it and this will be i'm just going to cut out some of the math and the network but this will give you a good overview of how you would proceed through the process first off we're going to look at the cd data if you take back to one of the the charts that jeff put up earlier we want to answer some of our preliminary questions do we have positive positive identification that the target is a lawful military target in this situation we're told yes we do under the current roe are we authorized to strike this type of target target in these conditions and we're told yes we're authorized to strike do we have a definitive boundary of the target yes we do and then we want to answer those last three questions which could be showstoppers is is it a dual use target is it an environmental hazard or is there a plume hazard and all three of those the answer is no to keep this thing to get to to keep this problem from getting too complicated so we're going to go look down and see the uh for the diagram what we have is a cluster of buildings and the blue rectangles are represent buildings the white rectangle with the triangle in the middle represents our target so we've got our collateral concerns a b and c and additionally some other buildings depending on what the radius of the expected effects are so we go down and we look on our chart and we see that all three of these buildings are single-story masonry buildings with uh target or collateral concern alpha and collateral concern bravo both 40 uh excuse me prior concern alpha and prior concern charlie both um 40 meters away from the target and collateral concern bravo 27 meters away from the target and the proposed weapon is a is a mark 83 our uh our effects radius for virtually all for for conventional weapons in the u.s arsenal and we see that if we were to drop the uh conventional weapon in the u.s arsenal with largest effects radius you're going to see a significant collateral concerns there so we're not going to be able to stop that collateral cd level one we're going to have to move on to cd level two and we know um we've got proposed pgm so level two and you'll see the circle there this accounts for the um for any pgm within the u.s arsenal and that radius we can still see that so we're not going to be able to start with we'll stop with cd level two analysis we'll have to move on to cd level three now you see when the circles lay down for cd level three and we uh look at just the effects of particularly a mark 83 either mitigated or unmitigated we're still going to have significant effects on collateral structures in the area so that tells us we're not going to be able to stop that at cd level three so move on to level four and this is where we take into account not only the mark 83 but the fact that the alpha bravo and charlie are all single-story masonry buildings so uh when you look at the collateral effects radius expected of a mark 83 versus a single-story masonry building we still have collateral uh collateral concerns we can still expect these collateral effects so what does that tell us it tells us we're going to have to move on to to cd level five now what do we need to know in order to conduct the cd5 analysis about these buildings we need to know first off what each of these buildings are used for and that's going to help us to find the appropriate values in the population density tables we need to know the size of the building so that we can calculate um how much of it's affected we need to know the percentage affected and we need to know the location in relation to the inner and the outer analysts of that effect range now for purposes of this i've uh i'm going to assume that we're going to have some values given to us by imagery analysts with regard to the size of the structures and the locations and the distances from the uh the impact and the percent effect so as you see on the um on the population density table that we put up the functionality of buildings alpha and charlie are both single family urban lower class with 2 000 square feet and you'll see that the imagery analysts have give us a percentage affected we take that percent affected and we multiply that we find we get our number of square feet affected of each of these buildings and then under our estimated population density and for the purposes of this i've made these numbers up for a couple reasons one because they vary from theater to theater to that because they're classified so these are these are made up numbers and you'll see that in alpha and charlie that we can expect four four people per thousand square feet six people per thousand square feet at night and then the episodic event is is inapplicable it's not a something that happens in a single family house so you'll see as the uh population as the um worksheet as you move from left to right on the worksheet we can expect that if you target alpha during the day you're going to have two uh you can expect two casualties at night you can expect three casualties and then an episodic event is so you look at um at collateral concern bravo as well now that this time it's a market so this is a little bit different still the same type of structure but it's got a different functionality so we can expect a different number of people in there during different times and episodic events we're told that it's about 2 000 square feet it's going to be about 10 affected so there's 200 square feet but in this particular region apparently markets aren't that populated during the week so only a couple people in there at any one time during the day at night maybe it's just the shopkeepers coming but in this particular area maybe everybody gathers on saturday morning for their weekly shopping so we have what's called an episodic event we can expect that on saturday mornings we're going to have 60 people show up at this market and be there for uh during that episode so you do the um you do the math and what you arrive at is if you attack this this structure you will cause one expected casualty during the day um but just just with regard to uh collateral concern a or excuse me collateral concern bravo one at night you'll notice that the math doesn't add up to a whole but we always round up any fraction is always rounded up to the next whole person but in an episodic event uh you wouldn't be expected to cause 12 collateral casualties so we added each of those we add those numbers up working from left to right to get the number for each uh each collateral concern and then add top top to bottom what we arrive at is if we attack this target with a mark 83 what we're going to what we can expect is during the day to have to cause uh seven seven casual collateral casualties at night to cause 10 collateral casualties and then 12 uh if it were to be struck during an episodic event so then you need to ask what else do i need to know in order to make a collateral damage estimate and what you need to know is that non-combatant casualty value what is the commander set as that level before it's got to be uh kicked up to a higher level so to speak and then for purposes of this problem we've got an ncv of eight so when you look at the the uh the total columns we can see that an episodic event or uh we're not going to work because we've got ncv of eight that will fly away that but however we are under the estimated casualties are under the ncv for a daytime strike so therefore we're going to have a casualty estimate of or excuse me a colonel damage estimate of a cd level five low required um restrictions are pgm mark 83 delayed fuse we've included a heading in there to avoid the uh to try to mitigate the weapons effects to the degree possible and then we also are required to include a caveat day attack only not during an episodic event in new market and what that's going to do is that that's going to leave us below that ncv of eight any questions on that at all or i guess i'm kind of curious i understand the methodology but it then gets compared to a value i guess my question is where does that value come from i mean that's got to be a methodology attached to how they got to the ncv because you're doing now a comparison the commander now does a comparison to come to now i think during you could have a varying level of ncv it's for instance in a coin environment there could be policy considerations that would drive that number down it could be uh depending on the stage of the operation um and i think it's it's reflective this doesn't do proportionality analysis for you but it assigns the uh it assigns the appropriate level commander to make the but what that what that does i think is reflects national policy as at what level do we need to raise we do we need to push this targeting decision up to the highest levels to make that proportionality analysis right the values would come from at least a geographic commander if not a higher commander as far as establishing what that limits are so it's basically at this you're looking at between a cd5 low and cd level five high and both are at pretty elevated levels of commanders that have to approve it but yes the five high would be the highest so it's going up potentially up to the national command authority level to approve so but yes that value would come from one of the those high level commanders so it's not something that's established down at the low tactical level and i also don't mean i don't believe it'd be anything that could be arrived at with any sort of mathematical equation driven by law and policy i think policy particularly as far as what what the command's willing what risk they're willing to assess to set and so depending on the mission right in a clean environment and it's probably pretty low you know international conflict it may be a higher value or different stages of the conflict all right uh if there's uh no other questions we'd like to thank everybody and uh just say once again this wasn't intended to make anybody cdu practitioners or experts in the area of collateral damage assessment or estimation but it was intended to give a a working knowledge for somebody that needed to that wants to discuss the product outputs to understand the process that went into getting that collateral damage estimation