U.S. aims to shoot down errant satellite

Remember the spy satellite that's falling out of orbit and is expected to hit Earth soon? Well, President Bush wants to shoot it out of the sky:

The Pentagon, under orders from President Bush, is planning to shoot down a broken spy satellite expected to hit the Earth in early March, the White House said Thursday.

U.S. officials said that the option preferred by the administration will be to fire a missile from a U.S. Navy cruiser, and shoot down the satellite before it enters Earth's atmosphere. ...

The disabled satellite is expected to hit the Earth the first week of March. Officials said the Navy would likely shoot it down before then, using a special missile modified for the task.

It's interesting that they want to shoot it down before re-entry -- which will create inevitable controversy, given that we condemned the Chinese for doing something similar last year -- when, according to satellite expert James Lewis, another option would be to try and use the Missile Defense System to destroy the satellite just as it begins to re-enter the atmosphere. That would be far less controversial (and would eliminate the "space junk" problem), but I'm guessing the odds of success are lower. In other words, reading the tea leaves, it certainly doesn't sound like the government has much faith in the Missile Defense System, at least as regards its ability to pull off a mission like this one.

Anyway...

 

Initially the administration believed that the danger from the falling satellite did not pose a large problem, but decided it was best to shoot it down when experts decided that the unused hydrazine did pose a danger.

That's the official line, anyway. But as the article also notes, "one of the main goals of the satellite's destruction is to prevent any sensitive equipment from falling into the wrong hands."

Here's some more on what would happen upon re-entry, if the satellite were allowed to re-enter.

It is not known where the satellite [would] hit. But officials familiar with the situation say about half of the 5,000-pound spacecraft [would be] expected to survive its blazing descent through the atmosphere and [would] scatter debris — some of it potentially hazardous — over several hundred miles. ...

Short-term exposure to hydrazine could cause coughing, irritated throat and lungs, convulsions, tremors or seizures, according to the Centers for Disease Control and Prevention.

Long-term exposure could damage the liver, kidney and reproductive organs.

Where the satellite would land would be difficult to predict until it descends to about 59 miles above the Earth and enters the atmosphere.

It would then begin to burn up, with flares visible from the ground, said Ted Molczan, a Canadian satellite tracker. From that point on, he said, it would take about 30 minutes to fall.

Comments

[Marty West]

http://www.youtube.com/watch?v=upenR6n7xWY

[Mad Max, Esquire]

Human safety is the reason the U.S. is giving for shooting it down. I think they want to vaporize it so they don't risk any of its technologies falling into the hands of the Chinese. Personally I think this is a totally acceptable thing to do.

[Mike]

Supposing this satellite wasn't shot down and then a large fragment dropped onto a kindergarten somewhere - you bet that those in the world who decry the US as trigger happy cowboys would be the first to ask questions about why this thing wasn't shot down when the US had the means to do it.
Re: space debris - the Chinese FengYun 1C satellite was attacked at >800km altitude. At that altitude the majority of the thousands of fragments created will last in orbit for decades, some for over a century as there is no atmospheric drag. In contrast this US satellite is already at an altitude of 265km where significant drag is being experienced. The object is dropping at over 0.8km/day.
Even if the thing is pulverised into thousands of pieces of shrapnel, the perigees of those new independently orbiting satellites will be very low. They'll be gone within weeks at the most (the majority will be down within hours), not decades.

[Sean Sullivan]

In practice, there is a major problem with intercepting the satellite (USA 193) as it begins re-entry: we don't know where it will happen.

The missile defense system is not designed as a worldwide shield (a la Brilliant Pebbles), but as a regional system for intercepting missiles over particular geographic regions. If the satellite happens to re-enter somewhere else (as is likely), the Missile Defense System would be useless.

For that reason, I believe that a space intercept makes much more sense (if you decide that you want to intercept at all, of course).

What makes me incredibly bothered is the likely inappropriate media and political comparisons between this and the Chinese ASAT test.

The problem with China was that debris (vast, vast, vast quantities of debris) was left in stable orbits, where it will present a hazard to other satellites for thousands of years.

In this case, with the intercept at 240 km, ALL of the debris will have a perigee (closest point to Earth in the orbit) of 240 km or lower -- and anything in that kind of an orbit is going to re-enter almost immediately due to atmospheric drag. Ergo, no meaningful debris from this event.

It's like the difference between shattering a windshield and dumping the shards into a garbage can (the proposed US intercept), and doing the same thing where the shards of glass are dumped all over a school playground, the sidewalk, and the adjacent highway (what China did). Not the same, methinks.

And it's going to peeve me, to no end, to see people saying "they're both shooting down satellites" and reducing it down to that one fact. Misses the point entirely.

Grump.

[Andrew]

Sean and Mike, you captured my thoughts exactly.

I should add that an uncontrolled, de-orbiting satellite is quite unlike an incoming ballistic missile in that atmospheric re-entry will immediately cause the satellite to fragment, burn, and re-orient to unpredictable trajectories and disparate velocities. A missile, meanwhile, has a fairly predictable re-entry pattern that theoretically can be tracked and targeted because its speed and trajectory, while perhaps not constant and unchanged by atmospheric drag and friction heat, are at least predictable. I do not know enough about the GMD program to comment on whether our deployed system is yet capable of intercepting an ICBM/warhead upon re-entry, but I can pretty much guarantee we don't yet have the ability to hit an out-of-control satellite once it re-enters Earth's atmosphere.

Once again, technological ignorance in the media drives a false public perception of both American missile defense capabilities and our policy judgment.

[Jim]

I think the concern is about debris being kicked up to higher orbits, which will inevitably happen.

Some interesting stats here:

http://www.armscontrolwonk.com/1795/the-shot

[Sandy Underpants]

Was someone arguing that the Chinese are more neat than Americans? They eat cats, rats and pigeons for crimney's sake! Not that there's anything wrong with that. I'm just sayin

[Sean Sullivan]

According to orbital mechanics, after any propulsive event (rocket firing, getting hit by a missile, etc), the range of the resulting orbit (perigee to apogee) will include the altitude of the spacecraft at the time of the propulsive event. Therefore, no debris can have a perigee higher than the impact altitude (240 km). From the perspective of long-term debris contamination of space, there should be zero concern about debris getting kicked up to higher orbits. Some will surely have a much higher apogee, but the 240 km perigee rule will ensure quick re-entry of all debris fragments. Nothing can remain in orbit.

There may be concern about the short-term effects of the debris: whether any of the fragments boosted to orbits with higher apogees can cause trouble during the brief time before their re-entry. That is an issue worth calculating, but space is really big, and without integrating the risk over a long period of time, the risk intuitively seems fairly small. I agree, though, that saying a risk is "an order of magnitude" less than a space shuttle flight is not too reassuring ...

Side note: I wrote my "school playground" example before reading Mike's "kindergarten" example, posted a few minutes earlier while I was writing. I guess the image of dumping hazardous substances on areas with lots of children is a good way to illustrate irresponsibility.

[4-7]

I suggest we first try to have a Humpback Whale (or at least a Manatee) talk the Satellite away. Shooting will only make it stronger.

[Jim]

4-7: I'm not sure I follow on your first paragraph. Saying that the range will include the altitude of the object at the time of impact borders on a tautology. It does not seem to imply, however, that debris cannot travel above and below this altitude.

I may be misunderstanding what you are trying to say. Do you have a reference tothe statement I might be able to read?

Based on the Chinese ASAT test, debris definitely was kicked up much higher. The Chinese satellite was destroyed at an altitude of about 537 miles.(1) Estimates indicate that the debris cloud reaches as high as 1,243 miles.(2)

I'm not sure if your assumption is based on the difference in altitude at time of impact, but it seems to me that if something explodes, it is going to send objects in all directions, regardless of how low its orbit is.

Like I said though, if you have something that could clear up any misconceptions I may have, I'd be happy to read it.

(1) http://www.fas.org/blog/ssp/2007/01/19
(2)http://www.aviationweek.com/aw/generic/story_generic.jsp?channel=awst&id=news/aw021207p2.xml

[Mike]

Jim - The FengYun 1C event did indeed lead to many pieces having raised apogees - some went very high indeed:

http://heavens-above.com/satinfo.asp?lat=55.75137&lng=37.62903&alt=0&loc=Unspecified&TZ=RFTm4&SatID=30457

or

http://tinyurl.com/23988t

However, all these pieces had orbits that crossed the orbit of the progenitor object's orbit - 850-880km. [note that the example above has a perigee lower than this 850-880km band].

So, by wrecking the spy sat at 240km you will get pieces with higher apogees, but the perigees will still be at 240km or lower.

A 240km perigee will ensure that atmospheric drag continues to act on the fragments even if the mean motion (orbits per day) is lowered from 16 to 15.

[Andrew]

Jim, I'm not a scientist, but after looking up "apogee" and "perigee", I understood Sean's point just fine. The point he was making was that there is a fixed formula for the perigee and apogee such that, if fragments of the satellite are pushed to a higher apogee, they necessarily end up on an orbit with a much lower perigee than 240km absent any sustained acceleration, meaning they won't stay up in space for long and will quickly lose orbit and burn up in the atmosphere.
So, if the debris really did fly as high as 1,200 miles above the Earth, the new orbit would become so elliptical that the debris would likely enter Earth's atmosphere within the first few revolutions around the planet. Yes, there is still a risk to satellites in that orbit range, but as Sean said, "...space is really big, and without integrating the risk over a long period of time, the risk intuitively seems fairly small."

[Andrew]

Okay, Mike's explanation is actually more technically correct than mine, but essentially we're saying the same thing. :-)

[Sean Sullivan]

Consider a satellite in orbit, not firing rockets or being hit by missiles, just going along doing its thing in peace. Over the course of a single orbit, the satellite's altitude will vary between a low point (perigee) and a high point (apogee, on the opposite side of Earth), and then back to perigee. On each following orbit, it will just repeat this cycle, up-down up-down up-down etc.

Now let's consider the USA 193 scenario. At first, we've got a quiet satellite in a roughly 240 km circular orbit. It gets whacked. Immediately afterwards, every piece of debris is in a new peaceful orbit, going up-down-up-down.

Consider an arbitrary piece of debris. The key point is that the up-down range of this piece of debris must include 240 km, because it's in a peaceful unperturbed orbit only a fraction of a second after impact. Yes, the debris might wind up going to a much higher altitude - but it must return, on every orbit, to 240 km (and possibly keep going lower, but that's not required).

So you can get stuff in a 240 x 3,000 km orbit, debris might sail really high. But you can't (for example) have a piece of debris wind up in a 350 km circular orbit, or 350 x 3,000 km. That's just not physically possible. Because right after impact, the debris is going along peacefully and unperturbed at (say) 241 km and sailing outward - and by definition, something at 241 km can't be in an orbit that never gets as low as 241 km, because it's already there, reducto ad absurdum. And as Mike noted, that altitude (even just at perigee) is sufficient to ensure a prompt re-entry.

The space shuttle offers an illustration of the same principle. When the space shuttle's main engines shut down (8 minutes after liftoff), the launch is over, but the shuttle isn't really in orbit. Main engine cutoff happens at about 110 km altitude, so the shuttle's orbit (right after the launch is over) can't have a perigee higher than 110 km - because the altitude of the last propulsive event (110 km) has to be in the perigee-apogee range of the orbit.

For a specific example, the current shuttle mission was in a 58 x 230 km orbit after main engine cutoff (source: Jonathan's Space Report). That perigee is basically immediate re-entry altitude, the shuttle can't stay there. For comparison, the "official definition of space" (see the basis for the X Prize standards) says it's altitudes over 100 km. So at the end of a launch, the shuttle isn't even in an orbit that will stay in space for a complete orbit!

The shuttle uses an orbital circularization burn about 40 minutes after launch to enter a real orbit. You have to wait until you coast to apogee before you can make a propulsive event that raises your perigee to make it circular. If that burn couldn't be done for whatever reason, the shuttle would make a landing about 90 minutes after launch, a situation called an "abort once around". That scenario is nearly certain to never happen, but it's interesting to note the importance of the firing a half-hour after the launch is "over".

Andrew, the formula you're speaking of relating apogee and perigee sounds like a mathematical relationship for computing one from the other (given the orbital period). Which in this case, won't be constant before and after impact.

[Jim]

Ah, I see what you are saying now. I think the original confusion was related to the blog I originally linked to using the language, "kicked up into a higher orbit" when really he should of wrote "kicked up into the path of a higher orbit," which is how I understood what was written.

But yeah, I understand the debris doesn't last long, but as has been stated before, an order of magnitude less than the shuttle is still uncomfortably high.

[Sean Sullivan]

Here's a back of the envelope calculation of the risk of another spacecraft colliding with a USA 193 fragment.

Let's consider a single spacecraft with an effective diameter of D meters, and a single piece of debris in an orbit with perigee P and apogee A km. We assume that the spacecraft is orbiting at an altitude between P and A.

We can assume that these objects will be in different orbital planes, since low-orbit spacecraft are almost never co-planar, except by design for related satellites in operational constellations. The only possibility of impact is on the line where the two planes intersect. Therefore, we will consider the "planar crossings" in isolation. This happens twice each orbit.

So let's calculate the risk to the satellite each time it passes through the orbital plane of the debris.

At a planar crossing, imagine a line directly away from the Earth extending from altitude P to A. To represent the satellite, at its altitude, highlight a length of diameter D. In order to get an impact, two things have to happen:

a) the orbit of the debris, at this point around the earth, has to pass through the highlighted region (else it will pass safely above or below the satellite on this particular orbit);

b) the debris has to actually be there (if it's on the other side of the earth at the time, then there will not be an impact).

With this model, I'll make some crude simplifications.

For (a), let's assume that the altitude of the debris at the planar crossing is simply chosen randomly from the allowed range P to A. Then the odds of the orbits crossing is D / (A-P) -- the ratio of the length of the highlighted area to the length of the whole line. Throw a dart blind at an area from here to there that includes a target, what are the odds you'll hit the target?

For (b), it matters how long it takes for the satellite's entire bulk to pass through the orbital plane of the debris. That depends on how much the two objects' orbital planes diverge. If the planes are perpendicular (best case), it's the length of time it takes the satellite to move by its own diameter (i.e. from the time the leading edge touches the perpendicular plane, to the time the trailing edge touches that plane). That's the ratio of D to the circumference of the orbit. (I assume the orbital periods of the satellite and debris are roughly similar, so saying "the duration of risk is 1 part in X of the satellite's orbit" leads to "1 chance in X of the debris being in the wrong place on this orbit".) If the planes are not perpendicular, the satellite won't clear the plane as quickly, so the debris can be farther away from the satellite's orbital plane when the center of the satellite crosses the planar crossing, and still have an impact with the edge of the satellite. For an oversimplification, let's just use the basic perpendicular case. The circumference of an orbit is roughly the circumference of the earth, so we can estimate this risk as (D / 40,000 km).

So we have the risk for a satellite, each time it crosses the orbital plane of the debris (twice per satellite orbit): [D / (A-P)] * [D / 40,000 km].

Units check: distance/distance * distance/distance = unitless number. Good. (Just make sure to keep consistent use of either m or km.)

Let's run this for ISS, and for a debris fragment that just barely reaches ISS at apogee. This is likely the worst-case scenario for any satellite because A-P will be small, and D will be big.

D = let's call it 50 m = 0.05 km
P = 240 km
A = 330 km

We get 1/1,800 * 1/800,000 = 1 / 1.44 billion

More typical numbers, for a debris fragment that just reaches a satellite in a 500 km orbit:

D = 0.01 km
P = 240 km
A = 500 km

We get 1/26,000 * 1/4 million = 1/100.4 billion

A satellite will have a planar crossing approx every 45 minutes (32/day), that's half the orbital period of around 90 minutes. Let's say the average relevant debris stays in orbit for three weeks. That gives 672 planar crossings to integrate the risk over.

For the ISS example, that's 1 in 2.1 million. For the more typical satellite, that's 1 in 149 million.

Let's say the USA 193 debris cloud produces 100,000 pieces. Most of that won't be kicked into an orbit with an apogee high enough to reach ISS, so let's say only 10,000 pieces are of possible impact concern, and most of those will have much higher apogees so A-P will be large. So on balance, let's say the debris risk is equivalent to about 2,000 pieces of debris of this worst-case (ISS orbit grazing) type. We'll use the 2,000 piece estimate for the higher satellite, too.

For ISS, this yields an overall risk of 1 in 1050. That's, uh, disturbingly close to an order of magnitude less than the risk of a shuttle flight. I wasn't expecting that, I thought I'd get a number that was much smaller! and in fact I started this exercise to show that the perception of risk from that remark was being blown up into more than was really there.

I don't really know about the characteristics of the debris generation, although I've read some about the FengYun debris field and looked through the orbits of catalogued pieces. So perhaps the main area of uncertainty in this estimate is the count of how many pieces might wind up in high enough apogee orbits, and how to translate the cumulative risk into the effective count of worst-case ISS-grazer examples used in these calculations.

For a more typical satellite in low orbit, the risk calculation yields 1 in 74,500.

[Andrew]

Brendan, where do you find all these really smart people from disparate backgrounds to read and comment on your blog? ;-)

[Mike]

If no action was taken against this satellite it would decay around March 20th according to Alan Pickup's SatEvo orbit propagator and decay predictor program.

The 100,000+ pieces of shrapnel/powder produced by the missile attack will all have a higher ballistic coefficient then the progenitor satellite so we can expect the overwhelming majority of pieces to be gone long before March 20th.

Of the pieces that get higher orbital apogees following the attack, of significance to any ISS threat is how the orbits of USA 193 and the ISS are orientated at attack time.

If I was being asked for my opinion about where to conduct the attack I'd recommend the northwest Pacific during a descending pass. Half-an-orbit later the debris cloud will be over the south Atlantic. Those pieces that lose orbital energy as a consequence of the attack will re-enter around there; the pieces that have their apogees raised will be climbing to or beyond the altitude of the ISS at that point. You wouldn't therefore want the ISS to be in the vicinity at that time.

Using SatEvo to propagate the ISS's orbit forward to March 20th I can see that the Right Ascension of the Ascending Node (RAAN) – the number that describes an orbit's orientation along with its inclination - of USA 193 (~352) and the ISS(~222) are nicely separated as the former object drops through the 240km-altitude mark. So, the ISS won't be in the crosshairs of the debris cloud on orbit 1. Each succeeding orbit should see debris re-enter and I would expect most (all?) of it to be already decayed from orbit before the cloud disperses around to the ISS orbital plane.

These numbers would mean that many of the 672 planar crossings in Sean's example would in fact be hazard free (in the earliest hours/days after the attack) therefore favourably skewing the odds against collision.

Of course, the very least I would expect the US to be doing prior to the attack is to run computer simulations with 100,000+ virtual debris pieces to see if there are any uncomfortably close virtual encounters with either the ISS or other spacecraft. If the answer is "yes" too often then I'd scrub the attack.