### Forum on Risks to the Public in Computers and Related Systems

ACM Committee on Computers and Public Policy, Peter G. Neumann, moderator

## Contents

Air Force Bombs Georgia
henry cox
The Geomagnetic Storm of 13 March 1989
Brian Randell
Tolerability of Risk
Martyn Thomas
More magnetic stripe woes
Joe Morris
Dive Computers revisited
Henry Cox
Info on RISKS (comp.risks)

### Air Force Bombs Georgia

henry cox <cox@rand.ee.mcgill.ca>
Thu, 18 May 89 10:49:24 EDT
[ From the Montreal Gazette, 12 May 1989]

US AIR FORCE PROBES WHETHER TRANSMITTERS CAUSED BOMB TO DROP

Atlanta - US Air Force investigators are examining whether
electromagnetic radiation from military transmitters may have caused an
F-16 jet to accidentally drop a bomb on rural West Georgia last week,
and Air Force official said yesterday.

The possibility of electromagnetic interference, however, is only one
of several potential causes the Air Force and Army is investigating,
said Dee Tait, an official at Moody Air Force Base where the F-16 is
stationed.  A final accident report won't be ready for 30 to 90 days,
she said.

No one was injured in the May 4 explosion, but the 230 kilogram bomb
ripped through a wooded area and has prompted a Georgia congressman to
call for a review of Air Force flight procedures state wide.

According to forces officials, the inadvertent bombing occurred when one
of four armed jets from the 247th Tactical Fighter Wing at Moody was
training last Thursday over Fort Benning's "Kilo Impact Area" in
Muscogee County.

The pilot of the plane, who has not been identified, tried to release a
bomb over the practice range, but it would not drop.  As the pilot
circled back over Marion County, the bomb fell and its 90 kg of
explosives shook windows of houses 900 metres away.

[ Short explanation of EMI causes deleted ]

It [ EMI ] has been attributed to navigation problems with the Army's
UH-60 "Black Hawk" Helicopter, which has been banned from flying near
100 transmitters worldwide.

In the case of the F-16, high levels of electromagnetic radiation can
accidentally detonate electro-explosive devices, or EEDs, that release
bombs, missiles and fuel tanks from the underside of the plane,
according to an Air Force {\it Explosive Safety Standards} manual obtained
by the Macon {\it Telegraph and News}.

The vulnerability of Air Force planes with EEDs has become and issue at
Robins Air Force Base near Warner Robins, Ga., where the Air Force has
been shutting down part of the high-powered PAVE PAWS radar station
every time and EED-equipped plane lands at the base.

The Air Force operates four PAVE PAWS facilities, which use radar
powerful enough to probe objects in space.  A current study at the
Robins base is examining the power of the pulsed radar beams from PAVE
PAWS and whether it disrupts ultra-sensitive electronic equipment on
aircraft.

The partial shutdowns preceded a March 1988 Air Force report that stated
"the high power contained in PAVE PAWS pulses may pose a danger to
elecro-explosive devices carried on military and commercial aircraft."

Tait confirmed that the F-16 [involved in the incident ] had been equipped
with EEDs, tiny explosive charges that release the shackles that hold
the bomb onto the jet.  "They are looking into that," she said.
However, she added, "the bomb-release mechanisms on F-16s are designed
to preclude electromagnetic interference."

Henry Cox (cox@pike.ee.mcgill.ca)



### The Geomagnetic Storm of 13 March 1989

Brian Randell <Brian.Randell@newcastle.ac.uk>
Tue, 9 May 89 18:51:04 BST
A colleague drew my attention to an article in Radio Communication
(Vol. 65, No. 5, May 1989), which made me realise belatedly just how
vulnerable we are to the effects of magnetic storms. Below I excerpt
from the article, without permission.

THE GEOMAGNETIC STORM OF 13 MARCH 1989

Ted Harris and David Kerridge, Geomagnetism Group, British Geological
Survey, 29 March 1989.

"The largest magnetic storm for 40 years started at 2am on 13 March
1989... The intensity of the storm was such that the aurora borealis
(northern lights), normally restricted to high latitudes, was seen
clearly in the south of England, and there were reports of
observations of the aurora in Italy and as far south as Jamaica.

"The rapid changes in the geomagnetic field during the storm induced
voltages in power lines, transoceanic cables, and telephone and cable
TV networks. In Quebec, transformers in the Canadian electricity
supply tripped, blacking out large areas of the Province and plunging
more than a million people into darkness. (No doubt with a
consequential blip in the birth-rate in nine months time!)

"Ionospheric disturbances caused disruption of radio communications
and resulted in the loss of TV reception in some areas. Satellite
communications were also affected - as were satellite orbits as the
increased ionospheric density produced extra drag.

"The increased radiation at high level created such potential hazards
that a Concorde airliner on a transatlantic route took a more
southerly flight path to avoid subjecting its passengers to radiation.
Astronauts aboard the the space shuttle Discovery' would have been
prevented from working outside the space craft because of the danger.
The shuttle mission was recalled a day earlier than planned because of
computer malfunctions which could have been caused by the storm.

"At sea-level, North Sea exploration companies reported that down-well'
compass readings of up to 12 degrees! A Norwegian geophysical exploration
company reported that all surveying has been halted after receiving warnings of
the storm and its severity from GRG. The director of operations reported that
two navigation systems used to fix the position of survey ships, which were in
agreement prior to the storm, were now diverging. GPS (Global positioning
system) satellites experienced increased drag which retarded their orbits so
much that positional accuracy at the Earth's surface was lost. ...

"Solar activity is likely to peak during 1990 (Solar Maximum), resulting in
more magnetic storms and a generally high level of magnetic activity over the
next two years at least."

Brian Randell, Computing Laboratory, University of Newcastle upon Tyne



### Tolerability of Risk

Martyn Thomas <mct@praxis.UUCP>
Wed, 17 May 89 14:55:43 BST
I strongly recommend the publication The Tolerability of Risk from Nuclear
Power Stations, Health and Safety Executive, Her Majesty's Stationery
Office, December 1987.  It contains a thorough discussion of the way in
which society perceives, and tolerates, risks from different sources.  It
also contains interesting UK actuarial statistics (...in Britain, a man of
20 has roughly a 1 in 1000 chance of dying in a year, for a man of 40 it is
1 in 500, at sixty, it is 1 in 50 for a man, 1 in 100 for a woman ...).

groups.
--
Martyn Thomas, Praxis plc, 20 Manvers Street, Bath BA1 1PX UK.
Tel:    +44-225-444700.   Email:   ...!uunet!mcvax!ukc!praxis!mct



### More magnetic stripe woes

jcmorris@mitre.arpa <Joe Morris>
Fri, 19 May 89 09:21:12 EDT
Quick background: the Washington area Metro subway system uses fare cards with
a magnetic strip on the back.  You buy a card of some particular value; it
is debited as it is used (the fares are distance-sensitive) and each time
you exit Metro the remaining value is recorded *and printed* on the farecard.
With this in mind, the following news article appeared in the 19 May issue
of the _Washington_Post_, p. C7 (as usual, without permission):

DASH Magnets and Farecards: A Fatal Attraction

Alexandria's DASH bus system [a suburban transit system] thought it was
promoting public transit Wednesday when it gave riders 2,500 refrigerator
magnets in honor of national "Transit Appreciation Day."

Funny thing, though, how the magnets apparently erased the value of an
unknown number of riders' Metro Farecards, officials said yesterday.

"We didn't do it intentionally, and definitely apologize to our passengers
for any inconvenience," said DASH General Manager Sandy Modell.

Metro officials said riders can obtain new cards by mailing the now
useless ones to Metro's treasurer's office [...].

The number of Farecards affected and the potential amount owed riders was
not known yesterday, DASH and Metro officials said.

The value of a Metro Farecard is magnetically encoded when the card is
purchased.  Cards are scanned electronically when passengers pass them
through the fare gates, which automatically deduct the trip fare.

Apparently, the small thin magnets, which fit in wallets and change
purses, erased the Farecards when they were stored together, said Metro
spokeswoman Beverly Silverberg.

"It happens all the time" when women carry purses with magnetic clasps
or riders carry other types of magnets, Silverberg said.  Modell warned
that magnets can have the same effect on automated teller machine cards
and some credit cards.

DASH bought the magnets, which included the DASH telephone number, from
the American Public Transit Association, which offered them to transit
agencies across the country, Modell said.  She did not know if any other
systems were similarly affected.

DASH passed out the magnets "as a token of appreciation" to riders,
Modell said.



### Dive Computers revisited

henry cox <cox@rand.ee.mcgill.ca>
Thu, 18 May 89 10:48:02 EDT
Some time ago several submissions dealing with the potential risks of
dive computers (which automatically monitor the nitrogen level in the
divers blood, and tell him when he must surface, etc.) appeared in
RISKS.  Since then, I acquired one myself.  My experience might be of
interest to others.

Soon after Christmas, two friends and I purchased three Oceanic
Datamaster II (a particular brand) dive computers.  At the same time, we
also purchased "Slimline" compasses, which were designed to fit into the
same console.  All of the units (three of three) were eventually
returned due to defects:

1) Due to electromagnetic interference, when the computer was ON, the
compass would point in whichever direction the console pointed - making
completely useless.

2) One unit was broken when shipped (or was broken during shipping), and
never worked at all.

3) Among other features, the computer was supposed to report the "dive
time remaining", based on air consumption and no-decompression nitrogen
levels - whichever is less.  On one unit, this was not recomputed
correctly - stuck on 29 minutes, even when there was no air left in the
tank it was connected to.  (Yep, I'll just sit here and hold my breath
for 29 minutes...)

4) The last unit appeared to work correctly when checked out in the
pool, but failed completely on its first real dive, giving no readings
at all.

All three units were returned and replaced with and upgraded model, the
Datamaster Sport - all three of which have worked properly to date.
The EMI problem was fixed with a redesigned console boot, which moved
the compass further away from the computer.

Certainly, some of these problems should have been caught and corrected
by the manufacturer (Oceanic USA, inc.) - particularly the EMI
interference on the compass, which would have been obvious to anyone who
turned the unit on and tried to use it.  The other problems MAY have
been caused by shipping, although I doubt it, as the cartons arrived
undamaged.  In any case, presumably the unit will be subjected to some
rough handling during use, and should be designed handle it.

I think that the real problem here was improper and incomplete testing
of the product before it was shipped out the door - potentially, a VERY
serious RISK, given the nature of the activity is designed for.

As has been stated many times before, computer readouts are no excuse to turn
off your brain, and it is not wise to rely on any one instrument.  In my
case, I dive with two complete sets of gauges (my old set plus my dive
computer), and I continue to work out nitrogen levels for myself.  Doing
otherwise would be very foolish.

Henry Cox

DISCLAIMER:  I have NO CONNECTION whatsoever with OCEANIC USA or any
other dive equipment manufacturer (except that I own some of their
equipment).  The opinions stated above are my own.  The events which
inspired them are also mine (unfortunately).



Report problems with the web pages to the maintainer