ANBACIS II COMPUTER DOWNWIND PREDICTION FEEDBACK

File: aaapm_01.txt
Page: 01
Total Pages: 2
SUBJECT:  ANBACIS II COMPUTER DOWNWIND PREDICTION FEEDBACK


MEMORANDUM FOR COLONEL PHILLIP
SUBJECT: ANBACIS II Feedback

I. 	After having run several scenarios on the ANBACIS II system,
I have noticed some very disturbing shortfalls within the system
or in the manner with which our tests scenarios are being run.

2. 	The program was reported to be based on the NBC 2 report
received from the field. You indicated that plots would be
computed based on the data received and that it would take
varying short amounts of time based on the product we requested.
As a result, we select the basic footprint to save time and we
have yet to receive a full map version that has been useable.

3. 	To date the majority of plots received have been those canned
versions already in the data bank which best approximate that
data contained in our test NBC 2.  I find this to be very
unsatisfactory for the level of. technology invested in this
system. A prime example was shown last night in which a reported
airburst scenario resulted in exactly the same footprint as a
near ground burst scenario submitted the day before. This would
indicate a level of effort not much more than that of a
simplified downwind predictor.  Is this the system response or
the system operator's selection in an effort to save time?

4. 	Time of response is turning out to be nearly three times that
reported to the field during fielding of the system. A basic
plot, reported to take 10 minutes, is taking 30 to 45 minutes.
One even took two hours. What is the problem?

5. 	One scenario that included four SCUDs, a reasonable attack
parameter, was reported as four separate footprints. I am
curious what will happen if we have a barrage of 60 artillery
munitions or 600 MRLS. I hope that you can adjust the program to
incorporate the full number of munitions into one footprint.

6. 	Another concern is the accuracy of the footprints or in
reality, our ability to believe/use the footprint copy we
receive. For example, we received four footprints the other
night for the same scenario and none of them could be matched to
the other. Some had incomplete footprints, i.e. incomplete
circle at point of detonation, unattached circles and downwind
vectors, and open ended footprints, etc. Some had missing data
which shortened the transmission page, a clue you told me to look
for, which in turn made the reference points unusable. When and
if we get into the real thing, there won't be time for
retransmissions or guessing if the plot is something we can use.

7.	Communications lines are not as clear as we both had hoped.
I find it interesting that when we call from here, the
transmission is very clear. However, when the call is initiated
from your end, we very seldom get a clear or complete
transmission. Could you check out the line on your end?