[fa.arms-d] Arms-Discussion Digest V3 #16

arms-d@ucbvax.ARPA (03/23/85)

From: Moderator <ARMS-D@MIT-MC.ARPA>

Arms-Discussion Digest Volume 3 : Issue 16
Today's Topics:

		Missile accuracy (2 msgs)
		Star Wars Thoughts (3 msgs)
		Correct units for "throwweight"
		Nuclear build-down, CPSR, etc ..
	
----------------------------------------------------------------------

Date: Wed, 20 Mar 85 08:43:27 pst
From: alice!wolit@UCB-VAX.ARPA
To: ARMS-D@MIT-MC.ARPA
Subject: missile accuracy

>     (How many ICBMs would fly more-or-less
>     North-South trajectories near the North Magnetic Pole in a real war?
>     Almost all of them.  How many have been tested for accuracy under
>     such conditions?  None.  No ICBM has ever been tested on a North-
>     South trajectory at all, never mind the North Magnetic Pole.  All
>     the test ranges, both American and Soviet, run East-West.)
> 
> I used to believe this too, but did you know that SLBM's are tested
> from all azimuths?  Therefore, it is reasonable to suppose that ICBM's
> will be at least as accurate as SLBMs in the N-S ICBM mode.

Even if missiles are tested over all azimuths, unless they are tested
over the actual trajectories they will follow (hardly an experiment
the other side is likely to permit), it will be impossible to gauge
completely the effect of gravitational and magnetic anomalies on their 
flight.

Given the stability resulting from mutual low confidence in missile
accuracy, it is disappointing (and a little surprising) that the issue
of an ICBM flight test ban has received so little attention from
superpower arms negotiators.  A country depending upon a generation of
missiles that have never been flight tested is not going to be
confident enough to initiate an attack;  a country facing a generation
of untested missiles might be more likely to ride out a suspected
attack rather than launch on warning.  It would be easy to verify
compliance with a flight test ban; much more so than verifying
compliance with current treaties on encryption of flight test data.

Jan Wolitzky, AT&T Bell Laboratories, Murray Hill, NJ; (201) 582-2998


------------------------------

Date: 20 Mar 85 11:11 PST
From: Ted Anderson <OTA@S1-A.ARPA>
Subject: Polar ICBM Trajectories 
To: armsd@MIT-MC.ARPA

I always assumed that "they" had very carefully watched polar satellites
which are in virtually identical trajectories to the ICBMs and calculated
the location and magnitude of all the gravitational anomolies.  I also wouldn't
be a it surprised if they had put up some polar satellites in low orbits
and just watched them as the slowly decayed.  The earth turning under them
as they go giving them an arbitrarily good look at the earths polar
gravitational, magnetic and atmospheric environment.  If the satellite
eventually deorbits and crashes into the SU you send an appology for the
regretable "accident".  They're watching it at least as carefully are you
are and are probably happy to get the same data.
	Cheers,
	Ted Anderson


------------------------------

Date: Wed, 20 Mar 85 11:45:09 PST
From: Charlie Crummer <crummer@AEROSPACE.ARPA>
To: ARMS-D@MIT-MC.ARPA
Subject:  Arms-Discussion Digest V3 #13, SDI Foolishness

Always keep in mind that technologically SDI is foolishness of the highest
magnitude. [Ref. Hans Bethe, Sidney Drell, Freeman Dyson, and others of the
top physicists in the world.]  There is hardly anyone who is actually working
on the technological problems of SDI that thinks that it will ever work.  I
even talked to someone who said that if I only knew some of the classified 
information that he knew, I would be even more convinced that SDI is
unworkable. (Are you listening, Lowell Wood?)

It has to be sold to people and it will be sold to these people, as if it
matters what they think, as a bargaining chip.  We have all heard from the
administration that there are no such things as "bargaining chips".  Unh!
Executive branch speak with forked tongue! (One of the deep and fundamental
facts of life.)

  --Charlie

------------------------------

Date: 20 Mar 85 16:08 EST
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  Star Wars Thoughts
To: CAULKINS@USC-ECL.ARPA
cc: ARMS-D@MIT-MC.ARPA

concerning the amount of code for SDI software systems: the SDIO has
now been quoted as saying that it will take between 10-100 million
lines, so the 6 M line seems conservative (see Washington Post Weekly,
National Edition, March 18, 1985).


    The cost-effectiveness of SWD is crucial - if the cost tradeoff
    between penetration and saturation methods and SWD doesn't indicate
    SWD a clear win, the Soviets will move in the less expensive and less
    risky direction.

Even if the cost-exchange ratio favors the U.S., the Soviets can
*still* overwhelm an SWD; all they need do is build missiles and hide
them from us.  During a war, we won't have time to replenish/maintain
our defenses.

    There are strong indications that SWD loses in any
    such comparison.  Several SWD countermeasures have been discussed:

    	Decoys - metallized balloons, some enclosing the real
    	warheads.

    	Chaff

    	IR emitting aerosols, and other things to screw up optical
    	sensors

    	Jamming and spoofing against sensors and communications links

You leave out the most significant one of all in this list, though you
return to it later -- the Soviet comparative advantage is in their
production capabilities and their warehouses.  They just haul out old
boosters and build cheap new ones (with or without guidance, warheads)
by the thousands.

------------------------------

Date: 1985 Mar 20 14:10:28 PST (=GMT-8hr)
From: Robert Elton Maas <REM@IMSSS.SU.EDU.ARPA>
To: Wedekind.es
CC: Arms-D@MIT-MC.ARPA
Subject: what are correct units for "throwweight"?
Reply-to: REM@MIT-MC.ARPA

> Date: 12 Mar 85 10:13:11 PST (Tuesday)
> From: Wedekind.es@Xerox
> Subject: US, USSR nuclear arsenal strengths
> 
> This table, bylined James Owens, appeared in Sunday's LA Times.
> ...
> * millions of pounds (sic).  Throwweight is the product of payload
> weight multiplied by the range of a weapon, and is one measure of the
> "destructive potential" of a weapon.

Obviously if you multiply weight by distance you don't get weight. I
presume you also divide by the number of launch vehicles or of
warheads (which?), to get unit throwweight instead of total arsenal
throwweight, right?

The correct unit must be something like:
  MegaPounds * Distance / ItemsCounted
where Distance is in feet or meters or miles or kilometers
and ItemsCounted is in launch vehicles or warheads.
Does somebody actually know what measure of distance is used and what
unit of ItemsCounted is used for ThrowWeight as used above?

------------------------------

Date: 1985 Mar 20 14:43:47 PST (=GMT-8hr)
From: Robert Elton Maas <REM@IMSSS.SU.EDU.ARPA>
To:CAULKINS@USC-ECL.ARPA
CC:Arms-D@MIT-MC.ARPA
Subject: BMD plan is destabilizing as designed, and useless in any case
Reply-to: REM@MIT-MC.ARPA

> Date: 19 Mar 85 07:08 PST
> From: CAULKINS@USC-ECL.ARPA
> Subject: Star Wars Thoughts
> To: ARMS-D@MIT-MC.ARPA
> 
> 6) There are hostile military systems against which current SWD
> designs will work better than against ballistic missiles; orbiting
> objects like parts of the opponent's SWD system, and surveillance
> satellites.  This kind of attack is clearly destabilizing, since its
> first effects are to reduce the quantity and accuracy of information
> available to the country under attack.  Also accidental weapon firing
> induced by software error is likely to escalate into a full-scale
> battle between the 2 opposing SWD systems.
> 
> The result of deployment of SWD systems by the US and USSR is more
> likely to lead to a new and unstable arms race in space than to some
> kind of defense-based peace.

This is precisely my argument against *any* kind of forward-basing of
defense systems. Forward basing is tantamount to offensive capability
(destroy the enemy's defenses) and thus destabilizing. I am therefore
opposed to boost-phase and coast-phase defense. Only terminal-phase
defense may be stabilizing, but at best can stop 80-90% of incoming
warheads if warheads are distributed uniformly. If warheads are
distributed non-uniformly, specific targets can be hit by saturating
the defense in specific areas. The result is that more warheads get
through but fewer targets are randomly penetrated, and since except
for particularily important targets the number of targets destroyed is
more than the number of warheads that reach target (in fact the latter
is almost totally meaningless), the optimal offensive strategy would
seem to be to distribute warheads more or less uniformly against
targets rather than to bunch them (except for particularily important
targets). Defensively, one should therefore harden important targets
(defense is meaningless since it can be penetrated by concentration of
enemy warheads) and otherwise defend all the rest of the targets uniformly.

In summary, Reagan's BMD plan has these major flaws as currently advertised:
 (1) It relies on three tiers (boost, coast, terminal), two of which
  are destabilizing thus undesirable. Hence all calculations of
  effectiveness are grossly overestimated compared to a stabilizing
  hence desirable terminal-phase-only defense.
 (2) Preliminary plans to defend only vital military targets are
  worthless since the USSR has plenty of warheads to penetrate such
  targets by concentrating warheads against them.
 (3) Ultimate plans to defend the whole of USA are worthless because
  enough warheads (10-20% of 25,000 warheads = 2500-5000 warheads)
  will reach target to destroy our society and create nuclear winter.
 (4) Of course the suspected secret plan of Reagan, to use BMD to
  soften a retaliatory strike by the USSR, so we can safely launch a
  first strike, is also destabilizing and disgusting for obvious reasons.

Further research may be desirable merely to refine our information and
explore alternative technologies, but faith in the system ending the
threat of thermonuclear anihilation in the forseeable future is foolish.

[Opinion of REM]

------------------------------

Date: Thu, 21 Mar 85 01:22:15 est
From: ericson@NYU-CSD1.ARPA (Lars Warren Ericson)
To: arms-d@MIT-MC.ARPA
Subject: Nuclear build-down/sharing technology

[]

Edward Teller and McGeorge Bundy "debated" last night on the virtues
of Star Wars in a public forum at NYU.  After an initial showing by the
Yippie Pie Thrower (at Teller; tofu and marinated mushrooms) and an attempt
to disrupt the meeting by some anti-meeting people, Teller propounded his
conception of Star Wars (as Bundy repeatedly pointed out, not equal to the
actual SDI proposal), going randomly into more specifics, after his usual
"I want to tell you, but it's secret", which was not going to wash because
Bundy was ready to go into specifics.

Bundy is patrician and acute as a debater; Teller was a little bombastic
and senile, but deep down, perhaps more pure of heart.  The debate was
stacked in the sense that both are hawks; for credentials, Bundy has
the Diem coup and the rest of Vietnam; Teller has the H-Bomb and a life of
opposition to arms control.  So it alternated between Teller saying we
should dig more fallout shelters and beware of the Russian menace, and
Bundy disparaging SDI deployment (but not research), because it hadn't
yet established itself as a cost-effective measure.

Two days ago I talked to a computer science professor here who is a member
of Computer Professionals for Social Responsibility (CPSR).  My 
understanding of that group was that it was started by computer scientists
in Cambridge and Palo Alto who refused to be paid by DoD/ARPA/ONR, and
that its most concrete activity was helping people find non-military
technical jobs.  With respect to Star Wars, he told me that they hoped
to function as a non-partisan, independent source of technical information
with respect to Government proposals.  When I suggested they could spend
their time promoting a specific policy position (such as nuclear build-down),
he said No, that was too partisan, they would lose their tax exemption.
(I read in the news yesterday that one member was filing suit against 
"Launch on Warning Capability", as a violation of the Constitution -- only
a President can declare war, and Presidents are not computers, ergo...)

Today I talked to a reputedly brilliant mathematics grad student, who told
me he was boycotting New Zealand lamb, (because of some variety of
farm-worker persecution Down Under...No...)  because they had not
"shouldered their part of the burden" of defending the Free World, when the
U.S.  nukes came to call.  When I offered the opinion that 1000 warheads
would be as militarily effective, and hence as credible a deterrent, as
10,000, that we could base those on land or in the air, and hence did not
need to lug them to New Zealand, he told me that only a *big* arsenal would
deter, that nukes are not supposed to be militarily effective, or they
wouldn't be a good deterrent, and that everybody should thus be willing to
have pieces of the arsenal in their backyard, so we can spread the feeling
of safety.

I am quite confused: perhaps others in ARMS-D can help me out.  It seems
that CPSR is quite partisan: is it not?  Are not 1000 missiles as effective
as 10,000, suitably modernized, diversely based?  Why did build-down bite
the dust?

As for SDI: Reagan Administration officials, at least a month ago, were
reported in the New York Times to have retracted the promise to share
information on SDI implementation with the Russians.  Why are people still
talking about this as if it were policy?  Given the Reagan Administration's
past history, this was promise was easily as specious as the identification
of ketchup with vegetables.

The promise itself, regardless of the utterer, has been attacked on two
grounds: first, the Russians could not trust our designs, because we could
be giving them a Trojan Horse; second, if we were not, they could turn our
openness against us, scrutinizing our true plans for weaknesses.

We could exchange sensor information.  Computer scientists could certainly
play a role in defining such an exchange, in the same way that engineers
defined the interfaces in the Apollo-Soyuz docking experiment.  Mutual
monitoring through the exchange of identical sensor data is not equivalent
to revealing defense architectures: it just makes both parties sure that
they agree on what is going up in the air.  Reliability of this exchange
could be improved through things like public-key electronic-signature and
encryption procedures.  A mutual identical sensor picture of global air-space
activity could go a long way towards reassuring people about the possibility
of accidental, machine-declared wars.

I think build-down and mutual monitoring are good policy ideas.  Why isn't
there a group of technical people willing to (partisanly) promote
constructive policy ideas, rather than run repeatedly into the opposing
brick walls of MAD ("liberals") and SDI ("conservatives")?

-- Lars Ericson
-- UUCP: ...!cmcl1!csd1!ericson
-- ARPA: ericson@nyu

------------------------------
[End of ARMS-D Digest]