[Back] [Index] [Next]

Operation Desert Storm:
Evaluation of the Air Campaign
(Letter Report, 06/12/97, GAO/NSIAD-97-134)


Appendix II

In this appendix, we respond to the requesters' questions about the
use, performance, and contributions of individual weapon systems used
in Desert Storm, particularly in regard to stealth technology and the
F-117.  We organize our discussion by four sets of subquestions, as

  Operating environment:  What predominant operating conditions
     prevailed during the air campaign?  Specifically, we examine the
     time available to the coalition to plan the air campaign and
     deploy forces to the region; the desert environment, the
     weather, and environmental factors that affected air operations;
     and the quality of the Iraqi threat, including Iraqi air defense
     capabilities and countermeasures to coalition bombing efforts. 

  Weapon system capability and actual use:  Based on original design
     or previous performance, what were the expected capabilities of
     the U.S.  air-to-ground aircraft and their munitions before the
     war?  Did performance during Desert Storm differ from
     expectations and, if so, in what way?  We assess patterns of
     aircraft and munition use during the war, such as the kind of
     targets to which aircraft were tasked; night versus day
     employment; the relative use of guided and unguided munitions;
     and the particular performance capabilities of the F-117.  We
     also evaluate official statements made before and after the war
     about the capabilities of aircraft and their respective target
     sensors in locating and identifying targets in various weather
     and when operating at night. 

  Combat operations support requirements:  What was required to
     support the air-to-ground aircraft in the form of refueling
     tankers, sensors, and suppression of Iraqi defenses?  We also
     address three controversies related to support for the F-117: 
     Did the F-117s receive radar jamming or other types of support? 
     What is the evidence that they were detectable by radar?  Did
     they achieve tactical surprise? 

  Survivability:  Were the survival rates of the various
     air-to-ground aircraft similar, and what factors affected
     aircraft survivability?  In particular, was the F-117 survival
     rate unique among these aircraft?  And were the defenses faced
     by the F-117s uniquely severe or comparable to those encountered
     by other aircraft? 

-------------------------------------------------------- Appendix II:1

In this section, we review the operating conditions in Desert Storm
with the object of distilling the lessons that can be learned for the

------------------------------------------------------ Appendix II:1.1

Following the Iraqi seizure of Kuwait, U.S.  forces had nearly 6
months to plan the air campaign and to deploy massive forces, many to
existing bases and facilities in Saudi Arabia and the other Persian
Gulf states, supplied in part from prepositioned stores as the
buildup proceeded.\1 The Iraqis chose not to interfere in any regard
with this massive buildup, leaving their own troops in static
positions as the coalition deployed increasingly large air, ground,
and sea forces.  The coalition had the luxury of time to deploy all
the forces it needed, along with their supplies, while the enemy did
little to obstruct the process.  In considering future contingencies,
and using Desert Storm as a baseline experience, it is important to
remember that the United States was permitted an uncurtailed buildup
of forces and military supplies to existing infrastructures on
foreign, yet friendly, soil that directly bordered the hostilities. 

The 6-month period also permitted identifying and studying important
strategic targets in Iraq.  Planners were able to extensively review
and revise plans for the critical strikes that took place in the
opening days of the air campaign.  During this period, many of the
units that saw some of the most activity in Desert Storm were able to
practice flying in the desert environment, honing their skills under
conditions for which some had not previously trained, given the
expectation that large-scale combat would most likely take place in a
European scenario.  There were opportunities to accumulate
intelligence on the nature of Iraqi defenses in part by intentionally
tripping Iraqi radars and observing Iraqi reactions.  In effect, the
U.S.  military services were able to plan their initial actions
thoroughly and in great detail, including the complex interactions
among dozens of U.S.  and allied military units, and to build up
large frontline forces and reserves without enemy interference. 

\1 See Operation Desert Storm:  Transportation and Distribution of
Equipment and Supplies in Southwest Asia (GAO/NSIAD-92-20, Dec.  26,

------------------------------------------------------ Appendix II:1.2

The vast, flat, open terrain of the KTO and Iraq was considerably
more favorable the effective employment of air power than most other
geographies around the globe.  While camouflage, gullies, and
revetments offered some possibilities for Iraqi concealment, almost
all analyses of the conflict conclude that, overall, it was easier to
find targets in the desert than in jungle or mountainous terrain. 
Moreover, until the ground campaign started after 40 days of air
bombardment, many Iraqi ground forces remained entrenched in fixed
positions, permitting repeated strikes against both personnel and

Cloud cover and storms made for the worst weather in that region for
at least 14 years, but conditions were no worse than what would
probably be the best ones likely in other conflicts.  At the same
time, because many air strikes were carried out at night, and some
under adverse weather conditions, the sensors used by aircraft and
munitions to locate, identify, and track targets were used under a
wide variety of environmental conditions. 

------------------------------------------------------ Appendix II:1.3

On paper, Iraq's air defense system appeared to be formidable to many
observers before the air campaign.  Iraq had purchased what was
widely described as a state-of-the-art integrated air defense system
(IADS) from France, which linked 17 intercept operations centers
(IOC) to four sector operations centers (SOC).  The IOCs were linked
to air bases with interceptor aircraft, as well as to dozens of
surface-to-air missile and antiaircraft artillery sites.  With
multiple and redundant communication modes, the system could, in
theory, rapidly detect attacking aircraft and direct antiaircraft
defenses against them.  (The IADS is described in app.  VI.)

However, the Iraqi IADS had been designed to counter limited threats
from either Israel, to its west, or Iran, to its east, not from the
south and north, nor from a massive coalition force to which the
United States alone contributed more than 1,000 combat aircraft.  As
the Navy's Strike Projection Evaluation and Anti-Air Research (SPEAR)
department reported before the war: 

     "the command elements of the Iraqi air defense organization (the
     .  .  .  interceptor force, the IADF [Iraqi Air Defense Force],
     as well as Army air defense) are unlikely to function well under
     the stress of a concerted air campaign."\2

Similarly, on almost every performance dimension, the Iraqi IADS was
remarkably vulnerable to massive and rapid degradation.  Evidence
from the Air Force, DIA, GWAPS, SPEAR, and other expert sources shows
that the principal deficiencies of the Iraqi IADS were that (1) it
could track only a limited number of threats, and it had very limited
capabilities against aircraft with a small radar cross-section, such
as the F-117; (2) its design was easy to disrupt, and the key IADS
nodes were easy to target, [DELETED]; and (3) many of its SAMs were
old or limited in capability, and the Iraqi air force played almost
no role in the conflict, although it had been intended to be a major
component of air defenses. 

In addition, the political context of the war permitted the
development of a strong, cohesive, coalition force while Iraq had few
allies, none of which were particularly strong or in a position to
materially aid Iraq. 

\2 Naval Intelligence Command, Navy Operational Intelligence Center,
SPEAR Department, Iraqi Threat to U.S.  Forces (Secret), December
1990, p.  3-14. 

------------------------------------------------------ Appendix II:1.4

Our review of unit after-action reports, lessons-learned reports, and
interviews with pilots suggests that Iraqi countermeasures to degrade
or impede the effectiveness of coalition air attacks or
communications were inconsistent and did not appear to have
represented as much as could have been achieved. 


Finally, toward the end of the war, the Iraqis ignited hundreds of
Kuwaiti oil wells, creating vast plumes of black oil-based smoke,
which seriously degraded visual observation and air reconnaissance as
well as the
infrared (IR) and electro-optical (EO) weapon sensors and the laser
designators on aircraft.  The purpose of this action appears to have
been more to punish Kuwait than to impede bombing efforts, although
it ultimately did this. 

It is difficult to assess the overall success of the Iraqi
countermeasures employed against aircraft sensors since it is not
readily known how many decoy targets were attacked or how many actual
targets were not attacked because they were effectively camouflaged
or hidden among their surroundings.  At the same time, given the
absence of attempted Iraqi jamming of satellite communications,
little if any jamming against coalition aircraft radars, and the
apparent absence of any discovery during or after the war that
countermeasures were used on a massive or even broad scale, it would
appear, on balance, that the use of countermeasures in Desert Storm
was inconsistent, at best, and did not seriously stress or impede
U.S.  aircraft sensors, bombing efforts, or communications. 

In sum, to answer our first subquestion, we found that a number of
unique political, logistic, intelligence, and threat conditions
characterized the environment in which Desert Storm took place. 
These conditions appear to have, at minimum, facilitated the overall
planning and execution of the air campaign and, therefore, must be
considered in assessments of Desert Storm outcomes and in
generalizing the lessons learned from this campaign. 

-------------------------------------------------------- Appendix II:2

The second major evaluation subquestion concerns the prewar
capabilities of air-to-ground aircraft, munitions, and sensors; their
stated prewar missions; and their actual use in Desert Storm.\3 In
this section, we discuss (1) comparing prewar aircraft mission
capabilities to actual mission use in Desert Storm, (2) examining
specific performance issues for the F-117, and (3) comparing prewar
target location and acquisition capabilities to capabilities observed
in Desert Storm. 

\3 A comparison of design and actual Desert Storm missions for
aircraft under review has the potential to reveal findings about the
attributes and limitations of the aircraft, the adequacy of pilot and
crew training, and the nature of the conflict.  For example,
deviations found between design and actual missions might reveal (1)
an inability of an aircraft to perform an expected mission, (2) an
unanticipated mission, or (3) a unique tactical environment. 

------------------------------------------------------ Appendix II:2.1

We compared official Air Force and Navy descriptions of the types of
combat missions for which their respective air-to-ground aircraft
were designed and produced to whether each aircraft actually
performed such missions in Desert Storm.\4 (See table II.1.)

                                    Table II.1
                     Air-to-Ground Combat Mission Categories
                      Attributed to Selected Aircraft Before
                        Desert Storm Versus Those Actually

ft        C\   DS\     C    DS     C    DS     C    DS     C    DS     \C     DS
------  ----  ----  ----  ----  ----  ----  ----  ----  ----  ----  -----  -----
F-117      X     X     N    \N     X     X     X     X     N     N    X\h      N
F-         X     X     N     N     X     X     X     X     N     N    X\i      N
F-15E      X     X     N     N     X     X     X     X     X     N      X      N
A-6E       X     X     X     X     X     X     X     X     N   N\j      X      X
F-16       X     X     X     X     X     X     X     X     X     N      X      X
F/A-       X     X     X     X     X     X     X     X     X     X      X      X
A-10     X\k     X     X     X     X     X     N     X     X     N    X\l      N
B-52       X     X     X     N     X     X     X     X     N     N      X      N
GR-        X     X     X     N     X     X     X     X     X     N      X      N

\a An "X" in column C (capability) indicates that the platform was
credited with the mission capability before Desert Storm (DS); an "N"
indicates that it was not credited with the capability.  An "X" in
column DS indicates that records show that the platform conducted
missions or strikes of this type in Desert Storm; an "N" indicates
that available records do not show this. 

\b Air interdiction (AI):  These are missions to destroy, neutralize,
or delay enemy ground or naval forces before they can operate against
friendly forces.  AI targets include transportation systems and
vehicles, military personnel and supplies, communication facilities,
tactical missiles, and infrastructure. 

\c Close air support (CAS):  These missions support ground operations
by destroying enemy capability in close proximity to friendly ground

\d Suppression of enemy air defenses:  These missions strive to
increase the survival or effectiveness of friendly aircraft
operations by destroying or neutralizing enemy air defenses. 

\e Offensive counterair:  These missions seek out and neutralize or
destroy enemy aerospace assets, such as airfields, aircraft in
shelters, and radar sites. 

\f Defensive counterair:  These are defensive air-to-air missions
flown against airborne enemy aircraft. 

\g Surface combat air patrol and joint maritime operations:  Surface
combat air patrol are sorties of naval aircraft to protect surface
ships from attack.  Joint maritime operations include the use of Air
Force aircraft to assist in the achievement of military objectives in
the naval environment. 

\h The F-117's JMO capability to attack naval targets at sea is
described as "minimal." It does, however, have the capability to
attack ships and other naval targets in port. 

\i Note h applies to the F-111F also. 

\j The A-6E is not credited with capability in this mission category. 
Only four DCA sorties were flown in Desert Storm; for that reason,
the cell has an "N."

\k The A-10's AI capability was described as limited in MCM 3-1 vol. 

\l The A-10's JMO capability was described as limited. 

Sources:  USAF TAC MCM 3-1 vols.  III, V, VI, XIII, XVII, XIX
(Secret), NAVAIR Tactical Manuals for the F-18 and A-6
(Confidential), official descriptions of the GR-1 from the Ministry
of Defense of the United Kingdom, and GWAPS, vol.  V, pt.  I
(Secret), pp.  336-404. 

We note in the table where the Air Force or Navy declared a mission
capability to be limited.  If an aircraft performed a very small
number of missions, such as fewer than five, we did not credit the
aircraft with exhibiting that capability in Desert Storm.  A very
small sample of missions does not permit the reliable determination
that the aircraft, successfully or unsuccessfully, demonstrated the

Table II.1 shows that in the four mission categories that emphasize
air-to-ground attack--AI, CAS, SEAD, and OCA--all the aircraft under
review were used to a meaningful extent during Desert Storm to
perform missions consonant with their stated capabilities.  In only
one case--that of A-10s carrying out OCA missions--was an aircraft
used for a mission for which it had not been envisioned.\5

The DCA mission category was one of two in which aircraft were not
used for a mission for which they had an acknowledged pre-Desert
Storm capability.  Except for F/A-18s, none of the aircraft under
review credited with a defensive air-to-air capability actually had
an opportunity to use it in Desert Storm.  Overall, nearly all of the
Iraqi aircraft that were shot down were attacked by F-15Cs. 

The relative paucity of air-to-air combat missions reflects the fact
that, for the most part, comparatively few Iraqi aircraft attempted
to attack either coalition aircraft or ground targets, despite the
fact that Iraq had about 860 combat aircraft and attack helicopters
combined.  Overall, the Iraqi air force essentially chose not to
challenge the coalition.  Over 100 Iraqi combat aircraft were flown
to Iran during the war. 

In sum, the data on intended versus actual Desert Storm mission use
indicate no substantial discrepancies between the anticipated
capabilities of aircraft and the missions for which they were
actually employed in Desert Storm.  Where stated capabilities were
not used, it was apparent that there was little need for them.  (See
app.  VII.)

\4 We excluded two types of missions that are highly
specialized--search and rescue and support of special operations

\5 Although Navy aircraft performed SCAP and JMO missions, Air Force
aircraft with this capability performed no significant number.  This
may have reflected a combination of sufficient Navy assets to deal
with these targets and traditional service rivalries. 

---------------------------------------------------- Appendix II:2.1.1

Our second evaluation subquestion further concerns whether the Desert
Storm data revealed particular patterns of aircraft and munitions
usage, on the weight of effort and type of effort measures, across
the 12 strategic target categories.  (See app.  I for a summary of
the WOE and TOE analysis.)

---------------------------------------------------- Appendix II:2.1.2

Many strategic targets were assigned basic encyclopedia numbers in
the target planning and study process.  Target assignment data that
include the number and type of aircraft and munitions were available
from the Missions database for 862 targets with BE numbers, including
kill box targets assigned individual BE numbers.\6 Figure II.1 shows
BE-numbered strategic targets in each of 12 categories that were
tasked to different types of aircraft.\7

The data in figure II.1 can be analyzed in terms of the pattern (or
lack thereof) in aircraft target assignments to BE-numbered targets
across the target categories, thus suggesting which aircraft, if any,
planners tended to prefer. 

In less than half the strategic target categories--that is, GVC, NAV,
NBC, SCU, and C\3 --did one or two types of aircraft strongly
predominate.  First, in the GVC category, F-117s were assigned to 27
(87 percent), F-16s to 8 (26 percent), and F-111Fs to 1 (3 percent)
of the BE-numbered targets.  Given that GVC targets were generally
high-value, in heavily defended areas, and sometimes either deeply
buried bunkers or heavily reinforced structures, the F-117's role
here appears consistent with its intended mission and the
capabilities of the specially designed warhead-penetrating I-2000
series LGBs with which it was equipped. 

   Figure II.1:  BE-Numbered
   Targets Assigned to Aircraft\ a

   (See figure in printed

\a The total BE-numbered targets depicted is greater than 862 because
some BEs were assigned to more than 1 type of aircraft. 

A preference pattern can also be found in F-117 assignments to NBC
targets (25 of 29, or 86 percent) and C\3 targets (151 of 229, or 66
percent).  In none of these was any other aircraft assigned to even
half the percentage accounted for by the F-117s.  However,
considerable redundancy among aircraft target assignments is
apparent:  while the F-117s were assigned to 86 percent of the NBC
BEs, the seven other aircraft, in sum, were assigned to over 90
percent of these BEs. 

Second, a strategic target category assignment preference was evident
in the NAV category, where two types of Navy aircraft, A-6Es and
F/A-18s, were respectively assigned to 83 and 79 percent of the 24
naval-related targets with BEs.\8

Third, a pattern of preference can be found in the SCU category,
where F-15Es were assigned to just over 68 percent of the 51
BE-numbered targets.  In contrast, the next highest participant
against these targets was the F-117, assigned to about 30 percent. 

Finally, in half of the strategic target categories--ELE, KBX, LOC,
MIB, OIL, OCA--no aircraft among those under review was alone
assigned to more than 60 percent of the targets or was otherwise
clearly predominant in terms of assigned BEs.\9 For example, in the
OCA category, all eight aircraft were assigned to between 27 and 48
percent of the BE-numbered targets, indicating very substantial
overlap among assigned aircraft and targets.  The data show similar
overlap in the five other categories (ELE, KBX, LOC, MIB, and OIL). 

In sum, the F-15E, F-117, A-6E, and F/A-18 were preferred platforms
against particular sets of strategic targets.  However, the general
patterns suggest that preferences, as revealed by patterns in target
assignments, were the exception and that among the aircraft reviewed,
most were assigned to multiple strategic targets across multiple
target categories. 

\6 KBX targets were mostly related to ground troops, for example,
tanks, artillery, and trucks located in large geographic areas.  (See
app.  I for a discussion of kill box targets.)

\7 This and similar analyses of the Missions database do not include
the A-10.  If the data on the over 8,000 A-10 sorties had been
usable, it obviously would have comprised a major part of these

\8 Clearly, 83 and 79 percent do not add to 100 percent.  When the
combined percentages of individual aircraft target assignments do not
add to 100, it means that at least two or more aircraft were assigned
to some of the same BE-numbered targets. 

\9 The F-16 was assigned to 51 percent of the BE-numbered KBX
targets.  However, a large number of the targets in this category had
no BEs assigned to them and are therefore not included in this
analysis.  Thus, the 51 percent for the F-16s may not most accurately
characterize the percentage of KBX-related targets that were assigned
to F-16s. 

---------------------------------------------------- Appendix II:2.1.3

Contrary to the general public's impression about the use of guided
munitions in Desert Storm, our analysis shows that approximately 95
percent of the total bombs delivered against strategic targets were
unguided; 5 percent were guided.  Unguided bombs accounted for over
90 percent of both total bombs and bomb tonnage.  Approximately 92
percent of the total tonnage was unguided, compared to 8 percent
guided.  These percentages characterized not only the overall effort
but also the proportion of guided and unguided tonnage delivered in
each week of the air campaign. 

Interviews with pilots and Desert Storm planners and a review of
relevant DOD reports, such as tactical manuals on aircraft and
munitions, identified reasons for this pattern.  Among these were (1)
poor weather and conflict-induced environmental conditions such as
smoke from bombing, which degraded or blocked the targeting sensors
required for the delivery of guided ordnance; (2) the comparatively
high cost of guided bombs and resulting smaller inventories (pilots
were frequently told to conserve guided bomb deliveries); and related
to inventory, (3) the fact that many strategic targets were large and
therefore generally appropriate for the use of unguided ordnance. 

The F-111F and the F-117 accounted for the majority of the guided
bomb tonnage delivered against strategic targets compared to the
other platforms reviewed.  Together, the 42 F-117s and 64 F-111Fs in
theater delivered at least 7.3 million pounds of guided bombs against
Desert Storm strategic targets over the course of the 43-day air
campaign.  Overall, more guided bomb tonnage was delivered against
OCA targets than against the other types of strategic targets, and
the F-111F accounted for the bulk of this delivery.  OCA targets
included hardened aircraft shelters and bunkers, which were
considered important and were targeted consistently, not least
because they housed much of Iraq's air force.  The achievement and
retention of air supremacy was critical to the successful, safe
continuation of the air campaign; thus, OCA targets were important. 

In at least one case--that of the Navy's night-capable A-6E--it
appears that capability to deliver LGBs was used only sparingly,
despite the fact that the 115 A-6Es deployed constituted almost 51
percent of all U.S.  LGB-capable aircraft on the first day of Desert
Storm.  A-6Es delivered fewer than 600 LGBs, or approximately 1.1
million pounds of bombs; these constituted about 7 percent of all the
LGBs used in the war. 

Summing across all target categories, the data show that, excluding
the A-10, F-16s and B-52s accounted for the preponderance (70
percent) of all unguided bomb tonnage delivered.  B-52s delivered at
least 25,000 tons (37 percent of total tonnage), and F-16s delivered
at least 21,000 tons of unguided ordnance against strategic targets
(31 percent).\10

\10 The tonnage delivered by A-10s is unknown but may have been
substantial given its sizable payload and more than 8,000 sorties
during the air campaign. 

------------------------------------------------------ Appendix II:2.2

Most strikes against strategic targets, including nearly all from
U.S.  LGB-capable aircraft, were conducted at night.  Five of the
eight air-to-ground aircraft under review carried out at least two
thirds of their strikes against strategic targets at night:  F-117
(100 percent), F-111F (99.6 percent), F-15E (94.2 percent), A-6E (72
percent), and B-52 (67 percent).  Figure II.2 compares the percentage
of day and night strikes. 

   Figure II.2:  Percent of Day
   and Night Strikes for Selected

   (See figure in printed

The use of the F-117 and F-111F nearly exclusively at night reflects
pre-Desert Storm expectations regarding mission capability.  Although
the F-111F can operate during the day, it has a designated emphasis
on night operations.  The F-117 can technically also operate during
the day.  But it was designed for night employment:  it is not
stealthy in day or low-light conditions, being readily visible to the
human eye.  Some of the design and performance characteristics that
make the F-117 low-observable to radar [DELETED] compared to other

The F-15E conducted 94.2 percent of its strikes at night, reflecting
a preference for this operational context since its stated mission
capability includes either day or night operations.  B-52s and A-6Es
also showed a preference for night operations, with more than two
thirds of their strikes against strategic targets conducted at night. 
Finally, the British Tornado was about evenly split on its percentage
of day and night strikes.  Overall, the data indicate that among the
air-to-ground platforms reviewed, more than half conducted two thirds
or more of their operations at night. 

The apparent preference for nighttime operations seems most likely
related to maximizing aircraft survivability.  As discussed later in
this appendix, in Desert Storm, optically guided Iraqi IR SAMs and
AAA were responsible for the largest number of aircraft casualties
(losses and damage).  Therefore, nighttime operations appear to have
enhanced aircraft survivability.  Further, in the desert environment,
the effectiveness of night attacks was improved for aircraft with
infrared targeting systems because operations at night provide
optimal heat contrast for some targets as the sand cools faster than
many objects in it. 

------------------------------------------------------ Appendix II:2.3

The F-117 has received highly favorable press for its achievements in
the Gulf War.  The Air Force has officially stated that the F-117
contributed much more to the Desert Storm strategic air campaign than
would have been expected given its limited numbers.  In its September
1991 white paper on Desert Storm, the Air Force stated that although
the F-117s made up only 2.5 percent of the aircraft in theater on the
first night of the war, they hit over 31 percent of the strategic
targets, and this pattern was exhibited both on the first night of
the campaign, when Iraqi air defenses were the strongest, and
throughout the remainder of the war.\11

Similarly, Lockheed, the primary contractor for the F-117, reported
that over the course of the war, F-117s represented only 2 percent of
total tactical assets yet accounted for 40 percent of all strategic
targets attacked.  The contribution of the F-117s was also
highlighted in DOD's title V report as the only aircraft to strike
targets in all 12 strategic categories. 

Clearly, the question of the relative contribution of the F-117, in
combination with claims about its accuracy (see app.  III) and
stealth characteristics, has important implications for future force
structure and procurement decisions.  In particular, we sought to
determine if the F-117 had been appropriately compared to aircraft
with similar missions and whether the data supported the claims made
for F-117 performance. 

\11 As recently as the February 1995 Annual Report to the President
and the Congress, the report of the Secretary of the Air Force stated
that "the F-117 destroyed 40 percent of all strategic targets while
flying only 2 percent of all strategic sorties during Desert Storm."
(See p.  300) While the portion of the coalition air forces
represented by the F-117 is addressed in this section, the accuracy
and effectiveness of the F-117 are addressed in appendix III. 

------------------------------------------------------ Appendix II:2.4

The 2.5 percent DOD cited as representing the percentage of F-117s in
the "shooter" force is derived from data that include many types of
aircraft that cannot bomb ground targets--the only mission of the
F-117.  Shooters are defined as aircraft that can deliver any kind of
munitions from bullets to bombs.  Table II.2 lists Desert Storm
combat aircraft classified as "shooters."

Not all shooter aircraft, however, can perform the same missions. 
Shooter aircraft include those that have solely air-to-air
capabilities as well as those that have air-to-ground capability. 
Since air-to-air shooters cannot hit ground targets but were included
in the shooter totals, the claim about the percentage of the total
shooter force that F-117s represented in Desert Storm is not
accurate.\12 Although they may have attacked 31 percent of the
strategic targets, they did not comprise only 2.5 percent of the
relevant shooters in the theater--that is, those that could deliver
munitions against ground targets. 

We sought to determine what percentage of the relevant aircraft they
did comprise.  On the first day of Desert Storm, 229 aircraft were
capable of both designating targets with lasers and autonomously
delivering LGBs.\13 The 36 F-117s in theater at the start of the
campaign were 15.7 percent of these 229 aircraft.  Thus, of all the
aircraft that had the potential to deliver some kind of LGB, the
stealth force represented not 2.5 percent of the assets but 15.7
percent.  Moreover, because the I-2000 series LGBs were only in the
Air Force's inventory, the F-117s actually constituted 32 percent of
all coalition aircraft that could deliver such bombs. 

                               Table II.2
                    Number and Percent of Coalition
                           "Shooter" Aircraft

Aircraft type                                           Number       t
------------------------------------------------------  ------  ------
F-117                                                       42     2.2
A-6E                                                       115     6.2
A-7E                                                        24     1.3
A-10                                                       132     7.0
AC-130                                                       8     0.4
AV-8B                                                       62     3.3
B-52                                                        66     3.5
EA-6B                                                       39     2.1
F-4G                                                        60     3.2
F-111E                                                      18     1.0
F-111F                                                      66     3.5
F-14                                                       100     5.3
F-15C                                                      124     6.6
F-15E                                                       48     2.6
F-16                                                       247    13.2
F/A-18                                                     169     9.0
A-4 (Kuwait)                                                19     1.0
CF-18 (Canada)                                              24     1.3
F-15 (Saudi Arabia)                                         81     4.3
F-16C/D (Bahrain)                                           12     0.6
F-5 (Bahrain)                                               12     0.6
F-5E/F (Saudi Arabia)                                       84     4.5
Hawks (Saudi Arabia)                                        30     1.6
Jaguar (France)                                             24     1.3
Jaguar (United Kingdom)                                     12     0.6
Mirage (United Arab Emirates)                               64     3.4
Mirage 2000 (France)                                        12     0.6
Mirage F-1 (France)                                         12     0.6
Mirage F-1 (Qatar)                                          12     0.6
Mirage F-1 (Kuwait)                                         15     0.8
Strikemaster (Saudi Arabia)                                 32     1.7
Tornado F3 (United Kingdom)                                 53     2.8
Tornado ADV (Italy)                                          9     0.5
Tornado ADV (Saudi Arabia)                                  48     2.6
Total                                                    1,875   100.0
Source:  DOD title V report, 1991. 

\12 The shooters total used to calculate the 2.5 percent figure
included not only air-to-air aircraft but also over 500 non-U.S. 
aircraft that never entered Iraq during Desert Storm.  Neither French
nor coalition Arab aircraft attacked targets in Iraq, although some
were used against Iraqi forces in Kuwait.  Thus, these coalition
aircraft did not represent aircraft that performed the same type of
mission as the F-117 (that is, attacking ground targets in Iraq). 

\13 Four types of LGB-capable aircraft and their respective
percentages in theater were 36 F-117 (15.7), 115 A-6E (50.2), 66
F-111F (28.8), and 12 F-15E (5.2).  Although the interdiction variant
of the Panavia Tornado, which the United Kingdom, Saudi Arabia, and
Italy had in theater, did deliver LGBs in a few instances, these
aircraft could not or did not autonomously operate with LGBs. 
Therefore, they are not included here.  Similarly, only the 12 F-15Es
that could autonomously deliver LGBs are included. 

------------------------------------------------------ Appendix II:2.5

Contrary to DOD claims, the F-117 represented approximately 16
percent of the Desert Storm LGB assets on day one and 32 percent of
LGB-capable aircraft that could deliver the penetrating I-2000 series
LGBs, particularly useful against hardened, reinforced, and buried
hardened targets.  Given this, it is not altogether surprising that
the F-117 seems to have been a preferred platform against GVC and NBC
targets.  The F-117 attacked approximately 78 percent of the targets
receiving LGBs on day one and attacked about one-third of all the
first-day targets, but it attacked less than 10 percent of all the
strategic targets that had been identified at the start of the air

During the first day of Desert Storm, F-117s performed 61 strikes,
which accounted for 57 percent of all first day LGB strikes against
strategic targets.\14 Three of the four LGB-capable carriers actually
delivered LGBs--the A-6Es, the F-111Fs, and the F-117s; F-15Es
delivered unguided munitions exclusively.  However, the F-117s and
F-111Fs accounted for all but about 7 percent of the strikes with
LGBs.  Fifty-nine BE-numbered targets received 108 strikes with LGBs. 
F-117 strikes represented 57 percent of these strikes (which were
against 46 of the 59 targets, or 78 percent); F-111F strikes were 36
percent of the total. 

\14 The first "day" was actually the first 29 hours in the Missions
database, from 1800 Zulu on January 16, 1991, to 2300 Zulu on January
17, 1991. 

------------------------------------------------------ Appendix II:2.6

One of the prominent claims the Air Force made for the F-117 in
comparing it to other bombers was that it, alone, attacked targets in
12 strategic target categories.  We found this claim to be accurate;
however, we also found that in three of the target categories--naval,
oil, and electricity--the F-117s attacked only one, two, and three
BE-numbered targets, respectively.  Further, we found that F-16s,
F/A-18s, and A-6Es each attacked targets in 11 of the 12 strategic
target categories; F-15Es attacked targets in 10 categories; and
B-52s and F-111Fs attacked targets in 9 categories.  As table II.3
shows, each of the other U.S.  air-to-ground aircraft in Desert Storm
attacked targets in no less than three-fourths of the target

                                              Table II.3
                               Coverage of Strategic Target Categories,
                                           by Aircraft Type

                                                                                         Tota  Perce
Aircraft            C\3   ELE   GVC   KBX   LOC   MIB   NAV   NBC  OCA  OIL   SAM   SCU     l     nt
-----------------  ----  ----  ----  ----  ----  ----  ----  ----  ---  ---  ----  ----  ----  -----
F-15E                 X     X    \a     X     X     X    \a     X    X    X     X     X    10     83
F-117                 X     X     X     X     X     X     X     X    X    X     X     X    12    100
F-16                  X     X     X     X     X     X    \a     X    X    X     X     X    11     92
F-111F                X    \a     X     X     X     X    \a     X    X   \a     X     X     9     75
F/A-18                X     X    \a     X     X     X     X     X    X    X     X     X    11     92
A-6E                  X     X    \a     X     X     X     X     X    X    X     X     X    11     92
B-52                  X     X    \a     X     X     X    \a     X    X    X    \a     X     9     75
GR-1                  X    \a    \a     X     X     X    \a    \a    X    X    \a     X     7     58
\a No targets in this category were attacked, by aircraft type. 

Although the F-117s attacked at least one target in each of the
12 categories, their taskings were concentrated on a narrow range of
target types within target categories.  These types of targets were
typically fixed, small, and greatly reinforced, being deeply buried
or protected by concrete.  F-117s conducted relatively few strikes in
categories where the targets were area or mobile (for example, MIB or
KBX targets).  Characteristic F-117 targets had known locations and
did not require searching. 

The relative contribution of the F-117 can also be assessed by
examining the number of targets assigned exclusively to it.  Table
II.4 shows that the F-117 was assigned exclusive responsibility for
more targets than any other aircraft among the 862 BE-numbered
targets for which there are data.  These targets were primarily in
C\3 , GVC, NBC, and SAM--categories that include known, fixed, often
hardened targets. 

                                              Table II.4
                               BE-Numbered Targets Assigned Exclusively
                                       to One Type of Aircraft

                                                                                  Unkno  Tota  Perce
Aircraft           C\3   ELE   GVC   LOC   MIB   NAV   NBC   OCA  OIL  SAM   SCU     wn     l     nt
----------------  ----  ----  ----  ----  ----  ----  ----  ----  ---  ---  ----  -----  ----  -----
A-6                  8     3     0     2     4     4     0     2    1    3     0      0    27   14.6
B-52                 3     4     0     0     8     0     0     2    2    0     0      0    19   11.7
FA-18                4     3     0     1     0     4     0     1    2    2     1      2    20   10.1
F-111F               0     0     0     6     0     0     1     6    0    1     1      1    16   13.6
F-117               94     3    13     7     7     0     8     4    2   27     3      7   175   46.3
F-15E               12     0     0     7     1     0     0     0    0    0    21      1    42   22.6
F-16                25     4     2     6     3     0     1     2    4    8     1      1    57   16.9
GR-1                 1     0     0     7     1     0     0     1    7    0     0     12    29   46.0
TLAM                 1     9     2     0     0     0     0     0    0    0     0      0    12   31.6
\a Percent of all target assignments that were exclusive. 

------------------------------------------------------ Appendix II:2.7

Here we address how the claimed prewar aircraft target acquisition
capabilities compared to those experienced in Desert Storm.  The
capabilities of aircraft to locate targets and then deliver munitions
accurately against them is intimately connected to sensors that aid
the pilots in carrying out these tasks. 

A series of steps must be performed to successfully attack a ground
target from the air, especially when precision munitions are being
used.  For fixed targets that have been previously identified and
located, the delivery aircraft must navigate to the geographic
coordinates of the target and then pick it out from other
possibilities, such as neighboring buildings or other objects.  For
mobile targets, the aircraft may have to search a broad area to find
and identify the right candidates for attack.  For either type of
target, the pilot may need to determine that the target is a valid
one--for example, the extent of previous damage, if any; for
vehicles, what kind; whether the object is a decoy; and so forth. 

------------------------------------------------------ Appendix II:2.8

Various sensor systems were used in Desert Storm to search for,
detect, and identify valid targets and to overcome impediments to
normal human vision, such as distance, light level (night versus
day), weather, clouds, fog, smoke, and dust.  These sensor systems
can be grouped into three technology categories:  infrared, radar,
and electro-optical.  (See app.  IX.) Each of these different sensor
technologies has been described to the Congress and to the public as
enhancing capability in poor visibility conditions, such as in the
day; at night; and in "poor," "adverse," or "all" weather conditions. 
Table II.5 shows the prewar official descriptions of the capabilities
of the sensors as well as their Desert Storm demonstrated

                                    Table II.5
                       Official Public Descriptions of the
                     Prewar and Desert Storm Capabilities of
                          Air-to-Ground Aircraft Sensors

                                  Prewar description of   Our findings on Desert
          Target search and       target-sensing          Storm actual
Aircraft  detection sensor        capability              capability
--------  ----------------------  ----------------------  ----------------------
F-117     Infrared (FLIR and      Night only;\b weather   Clear weather only;
          DLIR)\a                 is "a constraint not    flew exclusively at
                                  imposed by technology   night

F-15E     Infrared (LANTIRN)      Day and night;          All weather only with
          radar                   "adverse weather"       unguided bombs; clear
                                                          weather only for
                                                          guided munitions; flew
                                                          almost only at night

F-111F    Infrared (Pave Tack)    Day and night; "poor    All weather only with
          radar                   weather"                unguided bombs; clear
                                                          weather only for
                                                          guided munitions; flew
                                                          almost only at night

A-6E      Infrared (TRAM)\d       Day and night; "all     All weather only with
          radar                   weather"                unguided bombs; clear
                                                          weather only for
                                                          guided munitions; flew
                                                          day and night

F-16      Infrared, electro-      Day and night; "under   Clear weather only
          optical (LANTIRN) and   the weather"            (Maverick); all
          IR and EO               (LANTIRN); "adverse     weather only with
          (Maverick),\e and       weather" (Maverick)     unguided bombs; flew
          radar                                           day and night

F/A-      Infrared (FLIR) radar;  Day and night and       All weather only with
18\f      electro-optical         adverse weather         unguided bombs; clear
          (Walleye)               capability not          weather only for
                                  prominently stated      Walleye and FLIR pod;
                                                          flew day and night

A-10      Infrared and electro-   Day and night capable;  Clear weather only for
          optical (Maverick)      "adverse weather"       guided (Maverick) and
                                  (Maverick)              unguided munitions;
                                                          flew day and night

B-52      Radar                   Day and night and       All weather only with
                                  weather capability not  unguided bombs; flew
                                  prominently stated      day and night
\a Forward- and downward-looking infrared. 

\b Based on a postwar Air Force description; unofficial prewar
descriptions available to us did not make clear the night-only

\c Prewar unclassified descriptions were unclear about the F-117's
weather capability, so this is a postwar statement. 

\d Target recognition and attack multisensor. 

\e Some F-16s were equipped with LANTIRN navigation pods but no
targeting pods. 

\f See Naval Aviation:  The Navy Is Taking Actions to Improve the
Combat Capabilities of Its Tactical Aircraft (GAO/NSIAD-93-204, July

\15 Equipment and capabilities beyond those specifically described
and directly related to target sensing functions are not addressed. 
For example, separate navigation and air-to-air combat equipment and
capabilities are not assessed. 

------------------------------------------------------ Appendix II:2.9

Although desert environments are widely believed to exhibit
relatively nonhazy, dry weather providing uninhibited visibility,
there was actually great variation on this dimension in Desert Storm. 
Moreover, winter weather in the gulf region during Desert Storm was
the worst in 14 years.  Records show that there was at least
25-percent cloud cover on 31 of the war's 43 days, more than
50-percent cloud cover on 21 days, and more than 75 percent on 9
days.  Also, there were occasionally violent winds and heavy rains. 
As a result, the adverse-weather capabilities of the target-sensing
systems were frequently tested in the air campaign.  While the
frequency and severity of cloud cover and poor weather were not
comparable to more adverse weather conditions normal for other
climates, they were not nearly as benign as had been expected. 

IR, EO, and laser sensor systems demonstrated [DELETED] degradation
from adverse weather, such as clouds, rain, fog, and even haze and
humidity, [DELETED].  Sensors were also impeded by conflict-induced
conditions, such as dust and smoke from bombing.  In effect, these
systems were simply [DELETED] systems as characterized by DOD.  In
contrast, air-to-ground radar systems were not impeded by the weather
in Desert Storm.  This permitted their use for delivery of unguided
munitions, although usually with low target resolution. 

Similarly, night weapon delivery capabilities were tested, since as
noted previously, a large percentage of aircraft strikes were
conducted at night, including essentially all F-117 and F-111F
strikes and most F-15E strikes.  Of the more than 28,000 U.S.  combat
strikes and British Tornado strikes, about 13,000 (46 percent) were
flown at night. 

At the same time, a number of conditions during the air campaign
aided the effectiveness of target-sensing systems.  The flat, open,
terrain in the KTO, without significant foliage or sharp ground
contours, exposed targets to sensors and made all but the smallest
targets hard to conceal completely.\16 The desert climate provided a
strong heat contrast for targets on the desert floor, especially at
night.  The flat, monochrome nature of much of the terrain presented
a good optical contrast during much of the day for EO systems, by
making objects or their shadows--when camouflaged--salient.  The
Iraqi practice of deploying tanks in predictable patterns facilitated
their identification.  Similarly, because many Iraqi frontline ground
units remained in fixed positions for nearly 6 weeks of the air
campaign--essentially until the coalition ground offensive
began--they were easy to find and not difficult to distinguish from
friendly forces. 


\16 For example, there is evidence that the Iraqis took advantage of
areas where there was greater terrain variation to hide mobile Scud
launchers under bridges. 

---------------------------------------------------- Appendix II:2.9.1

Pilots generally reported that certain target sensors and bombing
systems gave them an effective capability to operate at night that
they otherwise would not have had.  These assessments were
particularly relevant to the IR sensing systems, such as LANTIRN, IR
Maverick, TRAM, Pave Tack, and FLIR/DLIR. 

[DELETED] F-15E pilots stated that they were "exponentially more
effective" with LANTIRN than without.  The A-10 was able to operate
at night in significant numbers [DELETED]. 

IR sensors proved important for effective night attack; however,
pilots of virtually every aircraft type also told us about a variety
of limitations. 

Effects of High-Altitude Releases on IR Sensor Resolution.  During
the air campaign, the majority of bombs were released from aircraft
flying above 12,000 to 15,000 feet because Brig.  Gen.  John M. 
Glosson ordered that restriction enforced after aircraft losses early
in the air campaign during low-altitude munition deliveries.\17
Higher altitudes provided a relative sanctuary from most air defenses
but resulted in a major compromise in terms of bomb accuracy and,
ultimately, effectiveness.\18 For example, some F/A-18 pilots
reported that bombing from high altitude sometimes meant a total
slant range to the target of 7 miles.  At this range, even large
targets, like aircraft hangars, were "tiny" and hard to recognize. 

Several methods were used to help overcome poor target image
resolution.  [DELETED]

Other Hindrances to IR Sensors.  Pilots reported that a variety of
environmental conditions, some natural and some conflict-induced,
impeded the capabilities of their IR sensor systems.  [DELETED]

\17 Brig.  Gen.  Glosson was Deputy Commander, Joint Task Force
Middle East, and Director of Campaign Plans for the air campaign. 

\18 In general, the higher an aircraft flew, the less vulnerable it
was to AAA, IR SAMs, and small arms fire. 

----------------------------------------------------- Appendix II:2.10

--------------------------------------------------- Appendix II:2.10.1


----------------------------------------------------- Appendix II:2.11

EO sensors depended on both light and optical contrast for target
searching and identification.  This obviated their use at night and
in any significantly adverse weather or visual conditions where the
line of sight to a target was obscured.  The requirement for visual
contrast between the target and its immediate surroundings imposed an
additional problem:  for Walleye delivery, F/A-18 pilots reported
that a target was sometimes indistinguishable from its own shadow. 
This made it difficult to reliably designate the actual target,
rather than its shadow, for a true weapon hit.  They also said that
the low-light conditions at dawn and dusk often provided insufficient
light for the required degree of optical contrast. 

F/A-18 pilots told us that a "haze penetrator" version of Walleye
used low-light optics to see through daytime haze and at dawn and
dusk, permitting use in some of the conditions in which other optical
systems were limited.  That notwithstanding, EO systems proved at
least as vulnerable to degradation as other sensors and lacked
full-time night capability. 

----------------------------------------------------- Appendix II:2.12


Despite the target discrimination limitations of most radar systems,
they had the advantage of not being impeded by adverse weather. 
However, even with this advantage, only comparatively inaccurate
unguided bombs could be delivered in poor weather since all the
guided munitions used in Desert Storm basically required clear
weather to enable their various IR, EO, and laser sensors and
designator systems to deliver munitions. 

-------------------------------------------------------- Appendix II:3

A realistic evaluation of the performance of combat aircraft in
Desert Storm involves acknowledgment of the nature and magnitude of
their support.  Here we address our third evaluation subquestion: 
What was required in Desert Storm to support various air-to-ground

Targeting activity and the success of strike aircraft are
inextricably linked to the performance and availability of external
support assets.  In many instances, aircraft relied on a number of
support assets to conduct missions:  for example, refueling tankers;
airborne control platforms like AWACS; airborne platforms that permit
battlefield command and control capability like JSTARS (Joint
Surveillance Target Attack Radar System); platforms that provide
fighter escort for strike aircraft (such as F-15Cs); airborne
platforms that conduct electronic warfare (such as F-4Gs, EA-6Bs, and
EF-111s); and airborne reconnaissance platforms that collect
intelligence and information used for BDA and those that detect and
monitor threats. 

Approximately 1,011 U.S.  fixed-wing combat aircraft were deployed to
Desert Storm, compared to 577 support aircraft, or a ratio of 1.75 to
1.\19 While combat aircraft outnumbered support aircraft in Desert
Storm, the latter flew more sorties--a fact that is important to
consider for future military contingencies.  Nearly 50,000 sorties
were conducted in support of approximately 40,000 combat
air-to-ground sorties, for a ratio of about 1.25 to 1.  Support
aircraft were relied upon for air-to-ground and air-to-air missions
in Desert Storm, both of which were conducted around the clock.  To
support the efforts of combat aircraft, the smaller number of
combat-support platforms would have had to fly more sorties. 

\19 See GWAPS, vol.  V, pt.  I (Secret), pp.  31-32.  Fixed-wing Air
Force, Navy, and Marine Corps aircraft as of February 1, 1991, are
the only aircraft included in the 1,011 total.  Aircraft identified
as "Special Operations" are not included.  Combat aircraft include
fighters, long-range bombers, attack aircraft, and gunships. 
Combat-support aircraft include tankers, airlift, reconnaissance,
surveillance, and electronic combat aircraft. 

------------------------------------------------------ Appendix II:3.1

---------------------------------------------------- Appendix II:3.1.1

One of the combat-support platforms that was perhaps most critical to
the execution of the air campaign was the aerial refueling tanker. 
Most Desert Storm combat missions required refueling because of
around-the-clock operations and the great distances from many
coalition aircraft bases and U.S.  aircraft carriers in the Red Sea
to targets in Iraq.\20 Virtually every type of strike and direct
combat support aircraft required air refueling.  At least 339 U.S. 
in-flight refueling tankers off-loaded more than 800 million pounds
of fuel.  For Air Force tankers alone, there were approximately
60,184 recorded refueling events.  On average, over the 43-day air
campaign, there were 1,399 refueling events per day, or approximately
58 per hour. 

Table II.6 shows the percentages of total known refueling events
accounted for by some of the U.S.  platforms reviewed here (data on
the F-117 were "not releasable").\21 Among all the known, recorded,
Desert Storm refueling events from U.S.  Air Force tankers, the F-16
and F-15 account for the highest percentages among the selected

                               Table II.6
                Percent of Total Known Refueling Events
                  for Selected Air-to-Ground Platforms

Platform                                                     Percent\a
--------------------------------------------------------  ------------
F-16                                                              23.0
F-15                                                              20.0
F/A-18                                                             9.5
A-10                                                               6.0
F-111                                                              4.3
B-52                                                               3.5
A-6                                                                3.4
F-117                                                               \b
\a Percentages of the total known number of Desert Storm refueling
events from U.S.  Air Force tankers only. 

\b Data were not available. 

To put the percentage of aircraft refueling events in context, we
examined the extent to which the number of known refueling events was
related to the number of strikes that platforms conducted.  We found
that the statistical correlation between the number of refueling
events and the number of strikes was large, indicating that among all
aircraft considered, there was a positive relationship.\23 In effect,
as the number of strikes conducted by all the included aircraft
increased, generally, so did the number of refuelings required by
those aircraft.  This is clearly illustrated by the F-16s, which
accounted for both the largest percentage of known aircraft refueling
events and the largest number of strikes among the platforms

\20 DOD's title V report (Secret), p.  115. 

\21 Although the number of F-117 refueling events was not available,
we developed an approximation measure in order to estimate a lower
bound of their number.  Based on the reported number of F-117 Desert
Storm sorties (1,299) and the minimum number of reported refueling
events per sortie (2), we estimate the lower bound of F-117 refueling
events to be 2,598, or 4.1 percent of a total of 62,782 from U.S. 
Air Force tankers only. 

\22 Not only U.S.  Air Force platforms received fuel from U.S.  Air
Force tankers.  Air Force tankers provided fuel for some non-Air
Force aircraft, including some Navy and Marine Corps aircraft. 
Therefore, the percentages reported in table II.6 are percentages
based on total number of refueling events for Air Force aircraft

\23 Pearson correlation coefficient, r = 0.69.  Strikes conducted
against strategic targets as reported in our WOE/TOE analysis, which
does not include F-117 data. 

---------------------------------------------------- Appendix II:3.1.2

In-flight refueling is a normal, routine part of air operations and
not one for which aircrew or tanker crew were unprepared.  However, a
number of factors in the Desert Storm environment caused this routine
process to become highly complex and sometimes quite dangerous for
tankers as well as other airborne platforms and, in instances,
resulted in restrictions or limitations in air operations. 

The use of large strike packages as well as constant,
around-the-clock air strikes resulted in heavily congested air space
during most of the air campaign.  The number of airborne aircraft was
sometimes constrained by the number of tankers that had to be present
to meet refueling needs.  Aircraft strikes on targets were sometimes
canceled or aborted because aircraft were unable to get to a tanker. 

To preserve tactical surprise as well as to keep tankers, which have
no self-protection capability, out of the range of Iraqi SAMs, nearly
all tanker tracks or orbits occurred in the limited airspace over
northern Saudi Arabia, south of the Iraqi border.\24 The heavily
saturated airspace alone increased the probability of near midair
collisions (NMAC).  Nighttime operations and operations in bad
weather only exacerbated an already complex, precarious, operational

The Air Force Inspection and Safety Center reported 37 Desert Storm
NMACs, believing, however, that these were only a fraction of the
actual number.  In one reported NMAC, a KC-135 tanker crew saw two
fighter aircraft approaching from the rear, appearing to be rejoining
on the tanker.  It became apparent to the tanker crew that the
fighters had not seen the tanker.  The tanker crew accelerated to
create spacing, avoiding an NMAC, but the reported distance between
the fighters and the tanker was only between 50 and 100 feet before
evasive action was taken. 

\24 We were told by several Desert Storm pilots, from different
units, that there were instances in which tankers had to cross over
into Iraq to refuel aircraft that would not have made it back to the
tanker before running out of fuel. 

\25 We made several recommendations for enhancing the efficiency of
aerial refueling operations based on Desert Storm.  See Operation
Desert Storm:  An Assessment of Aerial Refueling Operational
Efficiency (GAO/NSIAD-94-68, Nov.  15, 1993). 

------------------------------------------------------ Appendix II:3.2

The U.S.  air order of battle (AOB) during the third week of the air
campaign indicates that over 200 airborne sensor aircraft, providing
a range of combat-support duties, were in the Persian Gulf theater. 
These included a variety of reconnaissance, surveillance, electronic
combat, and battlefield command and control platforms.  A discussion
of the roles of each of these can be found in appendix X. 

------------------------------------------------------ Appendix II:3.3

Combat air patrol (CAP), escort missions, and SEAD are types of
combat-support missions that, in Desert Storm, were frequently tied
directly to aircraft strike missions or were conducted in areas near
where strikes were occurring and, therefore, also benefited strike

CAP missions protect air or ground forces from enemy air attack
within an essentially fixed geographic area.  In Desert Storm, these
included coalition ships, aircraft striking targets, and high-value
air assets such as AWACS and tankers.  Escort missions were normally
conducted by air-to-air fighter aircraft and were used to protect
strike aircraft from attack by enemy air forces en route to and
returning from missions.  In contrast to CAP, escorts do not remain
in a relatively fixed area but, rather, stay with the strike package. 
Fighter escort also served as force protection, when needed, for
airborne assets such as AWACS and tankers that have limited or no
self-protection capability.  Finally, jamming and SEAD support
aircraft like EF-111s, EA-6Bs, and F-4Gs provided direct support to
strike packages or target area support that benefited nearby strike

Figure II.3 compares the number of CAP, SEAD, and escort strike
support missions conducted during each week of the Desert Storm air
campaign.  Overall, the total number of CAP missions was somewhat
greater than SEAD missions and substantially greater than escort
missions, and there were no significant fluctuations in this number
during the 6-week air campaign.  That CAPs were often necessary for
combat-support aircraft (such as tankers and AWACS) as well as strike
aircraft may explain the greater number of CAP missions relative to
SEAD and escort missions.  In figure II.3, we also observe that the
only type of combat support-related activity that actually showed
some gradual decline over time was escort missions.  This is logical
given that the threat from enemy aircraft was significantly
diminished, if not eliminated, by the second week of the air

   Figure II.3:  Strike Support
   Missions by Week

   (See figure in printed

The only notable drop in SEAD missions was after the first week of
the air campaign.  However, the number remained rather static during
the following 5 weeks.  This may reflect the fact that although the
Iraqi IADS had been disrupted early in the air campaign, numerous SAM
and AAA sites remained a threat, with autonomous radars, until the
end of the war.  The fact that there was not a consistent decline in
SEAD missions, over time, suggests that simply destroying the
integrated capabilities of the air defense system did not,
unfortunately, eliminate its many component parts.  (This is
discussed further in app.  VI.)

------------------------------------------------------ Appendix II:3.4

The range of combat-related support encompasses some understanding of
the personnel required to maintain airborne assets.  In Desert Storm,
approximately 17,000 Air Force personnel had force maintenance
responsibilities.  This figure accounts for approximately 31 percent
of the total Air Force population in the area of responsibility. 

------------------------------------------------------ Appendix II:3.5

Shortly after Desert Storm, Air Force Gen.  John M.  Loh told the
Congress that

     "Stealth .  .  .  restores the critically important element of
     surprise to the conduct of all our air missions" and " .  .  . 
     stealth allows us to use our available force structure more
     efficiently because it allows us to attack more targets with
     fewer fighters and support aircraft."\26

In describing the performance of the F-117 in Desert Storm, another
Air Force general testified that

     "Stealth enabled us to gain surprise each and every day of the
     war.  .  .  .  Stealth allows operations without the full range
     of support assets required by non-stealthy aircraft."\27

In contrast, as discussed previously, conventional aircraft in Desert
Storm were routinely supported by SEAD, CAP, and escort aircraft. 
Because F-117s could attack with much less support than conventional
bombers, they were credited with being "force multipliers," allowing
a more efficient use of conventional attack and support assets.\28

For example, in their April 1991 post-Desert Storm testimony to the
Congress, Gens.  Horner and Glosson testified that 8 F-117s, needing
the support of only 2 tankers, could achieve the same results as a
package of 16 LGB-capable, nonstealth bombers that required 39
support aircraft or 32 non-LGB capable, nonstealth bombers that
required 43 support aircraft.\29 The Air Force depicted this
comparison in its congressional testimony with the graphic reproduced
as figure II.4.\30

   Figure II.4:  " The Value of

   (See figure in printed

   Source:  House Appropriations
   Subcommittee on Defense (Apr. 
   30, 1991), p.  472.

   (See figure in printed

In figure II.4, the use of the stealthy F-117 in Desert Storm is
depicted as having several positive effects:  it reduces the number
of aircraft employed on a mission, thereby reducing overall costs; it
reduces the number of aircraft and pilots at risk; and it increases
the number of missions that can be tasked without increasing the
number of aircraft.\31 However, following our review of after-action
reports and interviews with F-117 pilots and planners, we found that
this depiction does not adequately convey the (1) specific operating
procedures required by the F-117, (2) modifications in tactics during
the campaign to better achieve surprise, and (3) support, in addition
to tanking, that it received. 

\26 Testimony by Gen.  Loh (then USAF, Commander, Tactical Air
Command).  Department of Defense Appropriations for 1992, Hearings
before the Subcommittee on the Department, of Defense, House
Committee on Appropriations, Apr.  30, 1991, p.  510. 

\27 Testimony by Lt.  Gen.  Charles A.  Horner, then commander of 9th
Air Force and Central Command U.S.  Air Forces, before the House
Appropriations Subcommittee on Defense, Apr.  30, 1991, pp.  468-69. 

\28 Information that would definitively address the extent to which
the F-117s were detected by the Iraqis and the extent to which the
F-117s were supported by other airborne assets in Desert Storm is
classified.  We requested but were not granted access to information
that would have enabled us to fully understand the detectability of
the F-117 during Desert Storm.  Although that information could not
have been presented in this report, our review of it would have given
us greater confidence that the information contained in the report
was reliable and valid.  The information presented in this section
was the best we could obtain given our limited access to records. 

\29 House Appropriations Subcommittee on Defense (Apr.  30, 1991), p. 

\30 Figure II.4 depicts two actual strike packages employed against
the Baghdad Nuclear Research Facility.  Appendix XI addresses the
effectiveness of the conventional (F-16) and the stealth (F-117)
strike packages against this target. 

\31 The "value of stealth" depicted in figure II.4 is essentially
anecdotal--it depicts two missions flown during the first week of the
campaign.  The Air Force does not cite evidence that this represents
the typical, or average, use of support aircraft by conventional and
stealth aircraft in Desert Storm.  For example, because the standard
package illustrated for the conventional fighters was substantially
downsized by the end of the first week of the air campaign, as the
threat level was reduced, the claimed life-cycle cost for each of
these packages is not necessarily an appropriate measure for
comparison.  As discussed here, the depiction does not properly
credit other (nontanker) support assets that helped the F-117s attain
their Desert Storm achievements. 

---------------------------------------------------- Appendix II:3.5.1

In addition to its low observable features, the F-117 achieves
stealthy flight through the avoidance of daylight, active sensors or
communications, and enemy air defense radars. 

[DELETED] Every F-117 strike mission in Desert Storm was carried out
at night. 


Stealth Requires Extensive Mission Planning.  Each pilot has an
individual mission plan tailored to the assigned target and the
threats that surround the target.  Because F-117s are not "invisible"
to radar but, rather, as the Air Force points out, are "low
observable," a computerized mission planning system [DELETED]. 

\32 The F-117s were deployed to King Khalid Air Base near Khamis
Mushait in the southwestern corner of Saudi Arabia.  Mission times
averaged over 5 hours. 

---------------------------------------------------- Appendix II:3.5.2

A significant claim made by the Air Force is that because of stealth,
F-117s were able to achieve tactical surprise each night of the
campaign, including the first night when F-117s attacked the key
Iraqi air defense nodes and, in so doing, opened the way for attacks
by nonstealth aircraft, thereby greatly reducing potential losses. 
However, we found the following Desert Storm information to be
inconsistent with the Air Force claim. 

AAA Before and After F-117 Bomb Impacts.  A number of Air Force
officials told us that because AAA did not start until after the
first F-117 bombs had exploded, this was evidence that F-117s had
achieved tactical surprise.  However, we found that the absence of
AAA prior to bomb impact was neither consistent for all F-117
missions nor unique to F-117s. 

An Air Force after-action report stated that in the case of the A-10,
AAA began after the first bomb detonation, not just sometimes but "in
most cases" and in "the majority of first passes."\33

Similarly, pilots of other aircraft, including F-16s and F-15Es, also
reported the same phenomenon.  They encountered no AAA until after
their bombs exploded, and like the F-117s, they were subject to AAA
primarily during egress from the target.  Moreover, F-117 pilots told
us that, on occasion, AAA in a target area would erupt
"spontaneously"--before they had released their bombs or the bombs
had exploded.  In response to this threat, the F-117 Tactical
Employment manual states (on pp.  3-11, 3-29, and 3-31) that F-117
refueling and jamming support procedures were altered during Desert
Storm to delay "spontaneous" AAA in the target area. 


In sum, the claim that the F-117s consistently achieved tactical
surprise is not fully consistent with the information we obtained. 
The absence of AAA prior to F-117 bomb impact was not universally
observed and was not unique to the F-117.  [DELETED]

\33 57th Fighter Weapons Wing, Tactical Analysis Bulletin, Nellis Air
Force Base 92-2 (Secret), pp.  6-7 and 6-8. 

         F-117S BENEFIT FROM
---------------------------------------------------- Appendix II:3.5.3

In contrast to the Air Force illustration to the Congress that F-117s
require only tanker support in combat (see fig.  II.4), Desert Storm
reports and participants stated explicitly that the F-117s did, in
fact, receive more than just tanker support in Desert Storm. 

At the end of 1991, after press accounts stated that the Air Force
had exaggerated the degree to which F-117s operated without defense
suppression and jamming support, Air Force officials then concurred
that standoff jamming from EF-111s had been employed from time to
time in conjunction with F-117 strikes.\34

This position--that the F-117 did, in fact, benefit from jamming on
occasion--is more consistent with the title V report than with the
Air Force's testimony in April 1991 that failed to note nontanking
combat support having been provided to F-117s in Desert Storm.  As
discussed previously, the 37th Tactical Fighter Wing (TFW)
lessons-learned report unambiguously describes how jamming assets
were incorporated in F-117 tactics and operations.  Pilot interviews
and portions of the lessons-learned report also suggest that F-117s,
occasionally, benefited from fighter support aircraft. 


In terms of air-to-air fighter support, the Air Force states that
there was typically little or none provided for the F-117s.  The
Desert Storm "Lessons Learned" section of the F-117 Tactical
Employment manual is unclear on this issue, stating (on p.  3-29)

     "Unit coordination with the F-15s occurred each day.  While we
     never had any F-15s tied to us, we had to make sure they
     understood our general plan for the night."

In addition, several pilots we interviewed believed that air-to-air,
F-15 aircraft were in a position to challenge any Iraqi interceptors
that would have posed a threat to the F-117s. 

\34 Bruce B.  Auster, "The Myth of the Lone Gunslinger," U.S.  News
and World Report, November 18, 1991, p.  52, and Davis A.  Fulghhum,
"F-117 Pilots, Generals Tell Congress About Stealth's Value in Gulf
War," Aviation Week and Space Technology, May 6, 1991, pp.  66-67, as
reported in GWAPS, vol.  II, pt.  II (Secret), p.  354. 

-------------------------------------------------------- Appendix II:4

The percentage of aircraft lost and damaged in Desert Storm was very
low--compared both to planners' expectations and to historic
experience.  The attrition rates of the Israeli air force in the 1967
and 1973 Arab-Israeli wars were about 10 times those of Desert Storm. 

Coalition combat aircraft conducted approximately 65,000 combat
sorties in Desert Storm.  A total of 38 aircraft was lost to Iraqi
action, and
48 other aircraft were damaged in combat, making a total of 86 combat
casualties.  However, of these casualties, only 55 involved any of
8 air-to-ground U.S.  aircraft under review, of which just 16 were
losses, with the remaining 39 being damage incidents.  All coalition
aircraft casualties and the known causes are shown in table II.7,
with the aircraft under review listed first; for comparison, TLAM en
route losses are also shown.\35

                               Table II.7
                   Type of Coalition Aircraft Lost or
                      Damaged and Attributed Cause

Aircraft           SAM  IR SAM   AAA      Other            Total
--------------  ------  ------  ----  --------------  ================
F-117 lost           0       0     0        0                0
F-117 damaged        0       0     0        0                0
F-111F lost          0       0     0        0                0
F-111F damaged       0       0     3        0                3
F-15E lost           1       0     1        0                2
F-15E damaged        0       0     0        0                0
A-6E lost            1       0     2        0                3
A-6E damaged         0       0     3        2                5
O/A-10 lost          0       6     0        0                6
O/A-10 damaged       0       3    11        0                14
F-16 lost            2       0     1        0                3
F-16 damaged         1       2     0        1                4
F/A-18 lost          0       0     0       2\a               2
F/A-18 damaged       0       7     1        0                8
B-52 lost            0       0     0        0                0
B-52 damaged         2       1     2        0                5
GR-1 lost\b          4       1     2        2                9
GR-1 damaged\b       1       0     0        0                1
Other lost\c         2       6     3        2                13
Other                0       2     4        2                8
Total lost          10      13     9        6                38
Total damaged        4      15    24        5                48
Total               14      28    33        11               86
TLAM lost\d          0       0     0    [DELETED]        [DELETED]
\a One loss was attributed by GWAPS to a MIG-25; the second was
stated as unknown. 

\b GR-1 data in this table include aircraft from the United Kingdom,
Italy, and Saudi Arabia. 

\c These rows include AC-130, EF-111, F-4G, F-14, F-15C, AV-8B,
OV-10, A-4, F-5A, and Jaguar casualties.  While these aircraft are
not part of the focus of this report, they are included in this table
as part of our discussion of the effectiveness of the Iraqi air

\d TLAM losses are based on a study by Center for Naval Analyses
(CNA) and DIA that found that of the 230 TLAM Cs and D-Is, an
estimated [DELETED] did not arrive at their target areas.  An
additional 30 TLAM Cs with airburst mode warheads and 22 D-IIs could
not be assessed.  If the hit rate for these 52 TLAMS is assumed to be
the same as for the 230 assessable TLAM Cs and D-Is, then an
additional [DELETED] TLAMS did not arrive at their targets.  Thus, an
estimate for the total losses, using this assumption, would be a
minimum of [DELETED] and a maximum of [DELETED]. 

Source:  GWAPS, vol.  V, pt.  I (Secret), pp.  670-81. 

\35 By aircraft "casualties," we mean both aircraft that were lost
and aircraft that were damaged.  While some, but not all, damaged
aircraft were returned to service after repairs of varying extent and
while there can be important differences between an aircraft that is
lost and one that is damaged, we include damaged aircraft in our
analysis for the following reasons:  (1) air defense systems that
incur only damage nonetheless often achieve their aim of forcing the
damaged aircraft to return to base before the target is reached or
weapons are released; (2) DOD reports and statements made about
various aircraft refer not just to lost aircraft but also to hits
from air defense systems; and (3) including damaged aircraft is more
analytically conservative--that is, in assessing air defense systems
and aircraft survivability, it is impossible to predict for the
purposes of deriving "lessons learned" whether a hit will result in a
loss or merely damage. 

------------------------------------------------------ Appendix II:4.1

The system perceived before Desert Storm as most threatening--radar
SAMs--actually accounted for less than one-fifth the number of
casualties caused by AAA and IR SAMs.  Moreover, the system generally
considered to be a lesser threat, AAA, proved throughout the war to
be quite lethal. 

The data in table II.7 show that small, portable, shoulder-launched
SAMs with IR guidance systems were the leading cause of Desert Storm
aircraft kills, responsible for 13 of 38 (34 percent), followed by 10
(26 percent) attributed to radar SAMs and 9 (24 percent) to AAA.  In
contrast, AAA was the leading cause of damage to aircraft, accounting
for 24 of 48 cases (50 percent of total damaged).  IR SAMs were the
next leading cause of damage, with 15 cases (31 percent), and radar
SAMs were last, with 4 cases (8 percent). 

If we sum the losses and damage by cause, portable IR SAMs accounted
for 31 percent of the total casualties, and AAA accounted for 38
percent--both more than twice the 16 percent of total casualties from
radar SAMs.  In effect, the data show that the antiair threat
assessed by many both before and during the war as the "high" threat
system--radar SAMs--was responsible for just 16 percent of the
coalition's total casualties.  Conversely, the expected "low threat"
AAA and man-portable IR SAMs, such as the 1970's vintage SA-7, which
made up the majority of the Iraqi IR SAM force, accounted for 71
percent of total casualties (58 percent of total kills and 81 percent
of total damage cases). 

There are a number of possible explanations for this overall
inversion of the perceived high and low threats to combat aircraft. 
First, radar SAM sites proved vulnerable to attack and destruction
from U.S.  high-speed antiradiation missiles (HARM) and other SEAD
systems that were able to detect and thus locate radar systems and
directly attack them.\36 Every time a SAM radar was turned on, it
provided a beacon for the weapons that could attack it--as occurred
frequently, according to pilots. 

Second, and directly related, when the Iraqis operating the SAM sites
chose not to turn on their radars, to avoid being detected and
attacked, and then launched the SAMs ballistically--that is, without
radar guidance--the SAMs could not track a moving aircraft. 
Therefore, these SAMS had little, if any, chance of damaging
aircraft, which could easily evade them by maneuvering out of their

In effect, the radar that was critical to ensuring SAM lethality made
every SAM site vulnerable to destruction by U.S.  SEAD aircraft. 
Further, coordination among SAM sites was essentially precluded by
the fact that, as explained above, the Iraqi IADS proved vulnerable
to disruption and degradation very early in the air campaign.\37 As a
result, coalition aircraft were generally not threatened by a
well-integrated air defense system, with coordinated multiple defense
layers, but rather by hundreds of autonomously operating SAM and AAA
sites with individual radar(s), and by thousands of inherently
mobile, portable, shoulder-launched IR SAMs and thousands of AAA guns
without radars. 

Figure II.5 shows the day-by-day coalition aircraft casualties from
radar-guided SAMs for the 43 days of the war.  After day 5, aircraft
casualties from radar-guided SAMs dropped off sharply:  there were
nine casualties over the first 5 days but only five more from
radar-guided SAMs during the remaining 38 days of the war. 

In sharp contrast to the readily detectable and locatable
radar-guided SAMs (of which there were hundreds), neither IR SAMs nor
optically aimed AAA emit any signal during their search and
acquisition phase.  Moreover, there were thousands of AAA sites
throughout Iraq and the KTO and thousands of portable IR SAMs in the
KTO.  Except for the small number of fixed AAA sites that had, and
actually used, radar, all IR SAMs and most AAA were very hard to find
before they were actually used.  As a result, even at the end of the
war, pilots reported little if any diminution of AAA, and aircraft
casualties from AAA and IR SAMs continued up to February 27--at the
end of the war.  As the Desert Storm "Lessons Learned" section of the
F-117 Tactical Employment manual reported (on p.  3-29), "The threats
[to aircraft] were never attrited .  .  .  AAA tended to be the
highest threat."

   Figure II.5:  Combat Aircraft
   Casualties From Radar SAMS

   (See figure in printed

Note:  Air tasking order day (ATODAY). 

Figure II.6 shows clearly that 17 aircraft casualties occurred within
the first 24 hours, or nearly 20 percent of the war's entire aircraft
casualties (during less than 2.5 percent of its total length).  It
was during this time that Iraqi defenses were at their strongest and
were first attacked and that coalition pilots were at their lowest
levels of Desert Storm combat experience.  Similarly, there was a
significantly higher overall daily casualty rate in the first 5 days
of the war, during which 31 aircraft casualties occurred (36 percent
of the total and an average of 6.2 per day), compared to the
following 38 days, with a total of 55 more casualties (an average of
1.45 per day). 

This diminution in aircraft casualty rates may partly be explained by
the fact that losses to radar-guided SAMs fell to nearly 0 after day
5, having accounted for 29 percent (9 out of 31) of total casualties
by then.  They accounted for just 9 percent (5 out of 55) of all
aircraft casualties in the remainder of the war.  It is apparent,
therefore, that by the end of day 5 of the air campaign, radar SAMs
had been virtually eliminated as an effective threat to coalition

   Figure II.6:  Daytime Combat
   Aircraft Casualties From All

   (See figure in printed

Moreover, in the first 3 days of the war, some aircraft (B-52s,
A-6Es, GR-1s, and F-111Fs) attacked at very low altitude, where they
found they were vulnerable to low-altitude defenses--AAA and IR SAMs. 
As a result, on day two, Brig.  Gen.  Glosson ordered that all
coalition aircraft observe a minimum attack level of about 12,000
feet.  While probably improving overall survivability, this tactic
also resulted in much less accuracy with unguided weapons (see
discussion in app.  III).  In effect, Brig.  Gen.  Glosson's order
served to manage the attrition rate of the air campaign, taking into
account the view, as one general stated, that no Iraqi target was so
important as to justify the loss of a pilot's life. 

Since the effects of having degraded the Iraqi IADS cannot be easily
separated out from the effects of also consistently flying only at
higher altitudes, the extent to which the latter decreased
vulnerability cannot be quantitatively specified.  However, there are
data on the altitude at which 32 U.S.  Air Force aircraft casualties
occurred (data were not available for other aircraft).  Of these 32
cases, 21, or about two-thirds, were hit at or below 12,000 feet.\38
This suggests that the altitude floor did serve to save lives.\39

Figure II.6 also shows that after the first week, aircraft casualties
occurred sporadically, but there were 17 hits during the last week of
the war.  Since only two of these were attributed to a radar-guided
SAM, it is apparent that AAA and IR SAMs remained potential threats
to the end.  The casualty data therefore confirm the statements of
numerous pilots who told us that, unlike radar SAMs, AAA and IR SAMs
were never effectively suppressed, thereby continuing as lethal
threats throughout the war. 

\36 Aircraft with HARMs or those that engaged in SEAD included the
A-10, F/A-18, F-16, F-15E, F-117, F-111F, B-52, GR-1, F-4G, A-6E,
EA-6B, and EF-111. 

\37 See the "Operating Conditions" section above and appendix VI. 
See also Joint Electronic Warfare Center (JEWC), Proud Flame
Predictive Analysis for Iraq (Secret), San Antonio, September 1990,
p.  28. 

\38 Of those 21, 12 were A-10 casualties.  A-10s were permitted to
operate below 12,000 feet to as low as 4,000 to 7,000 feet on January
31 and thereafter.  After January 31 is when 10 of the 12 medium- to
low-altitude casualties occurred. 

\39 Additional evidence that low-altitude deliveries were more lethal
than higher ones can be found in the pattern of A-6E and British
Tornado losses.  Of the seven British Tornados that were lost, four
were shot down during the first week of the campaign at very low
altitude while conducting strikes against airfields.  In an analysis,
DIA concluded that the basic cause was delivering ordnance at very
low altitude in the face of very heavy defenses, rather than being
the function of a defect in the aircraft.  After the change to
medium-altitude deliveries, only three more British Tornados were
lost in the remaining 5 weeks of the air campaign.  A-6E pilots told
us that their casualty rate dropped significantly after units using
low-altitude tactics switched to high altitudes. 

------------------------------------------------------ Appendix II:4.2

Aircraft casualty rates can be calculated by dividing casualties by
total sorties or total strikes.  Table II.8 shows aircraft casualty
rates per strike for the aircraft under review. 

The overall aircraft casualty rate was 0.0017 per strike, or in other
words, about 0.0017 aircraft were lost or damaged per strike in
Desert Storm.  The F-117 was the only aircraft under review that
reported no losses or damage.  However, using an analysis performed
in DOD but not publicly reported, we calculated the likelihood of a
nonstealthy aircraft being hit if it flew the same number of strikes
as the F-117 (that is, 1,788), with a general probability of hit
equal to 0.0017.\40 This calculation showed that 0 hits would be the
most likely outcome for a nonstealthy aircraft conducting 1,788
strikes.  This indicates that although there were no F-117 casualties
in Desert Storm, the difference between its survivability and other
aircraft may arise from its smaller number of strikes as much as
other factors. 

                               Table II.8
                  Desert Storm Aircraft Casualty Rates

                                 Total       Total   Aircraft casualty
Aircraft                    casualties     strikes     rate per strike
--------------------------  ----------  ----------  ------------------
F-117                                0       1,788                   0
F-111F                               3       2,802              0.0011
F-15E                                2       2,124              0.0009
A-6E                                 8       2,617              0.0031
O/A-10                              20     8,640\a              0.0023
F-16                                 7      11,698              0.0006
F/A-18                              10       4,551              0.0022
B-52                                 5       1,706              0.0029
GR-1                                10       1,317              0.0076
Total                             65\b      37,243              0.0017
\a Precise A-10 strike data were not available.  GWAPS recorded 8,640
A-10 sorties.  Given the definition of a strike, the number of A-10
strikes may have been larger than the number of bombing sorties.  If
the number of A-10 strikes is larger than 8,640, then its per-strike
aircraft casualty rate would be lower. 

\b Totals do not conform to the total shown for all coalition
aircraft in this table because only the air-to-ground aircraft under
review are included. 

\40 This analysis considers only the number of strikes flown. 
Factors known to be related to aircraft survivability--for example,
the severity of defenses and the time of day when strikes were
conducted--were not factored into the analysis. 

------------------------------------------------------ Appendix II:4.3

Other ways to compare Desert Storm aircraft casualty rates put the
F-117's survival rate in a clearer perspective.  Since the F-117s
attacked only at night, we examined the casualties for other aircraft
during night missions, in effect controlling for daylight (when
optically aimed antiaircraft weapons can be used most effectively). 
Data on whether aircraft casualties occurred in day or at night were
provided for 61 of the 86 coalition aircraft casualties.  Twenty-five
(29 percent) were not identified as either day or night and were
presumably unknown or unrecorded.  Of the 61, 44 (72 percent) of the
casualties with a known time occurred in daytime; 17 (28 percent)
occurred at night.  (See table II.9.) These and other data strongly
suggest that flying combat operations at night was safer than flying
during the day. 

                               Table II.9
                  Aircraft Casualties in Day and Night

Aircraft             Lost          Damaged       Lost          Damaged
--------------  --------------  ----------  --------------  ----------
F-117                 \a                \a        \a                \a
F-111F                \a                \a        \a                 3
F-15E                 \a                \a        2                 \a
A-6E                  2                  1        1                 \a
O/A-10                6                 12        \a                \a
F-16                  3                  3        \a                \a
B-52                  \a                 2        \a                 1
F/A-18                1                 \a        1                 \a
GR-1                  3                 \a        4                 \a
Other                 11                \a        4                  1
Total                 26                18        12                 5
\a No casualties or no day or night data on casualties. 

Five types of aircraft--F-111Fs, F-15Es, A-6Es, A-10s, and
F-16s--flew at least as many night strikes as the F-117.  As shown in
table II.9, of these aircraft, F-111Fs, A-10s, and F-16s also
incurred no losses at night, and the A-6Es, A-10s, and F-16s received
no damage at night.  In this context, it is notable that the aircraft
that incurred the highest absolute number of casualties, but not the
highest attrition rate, the A-10, incurred neither losses nor damage
at night, although it conducted approximately the same number of
night sorties as the F-117.  These data suggest that, in Desert
Storm, flying at night was much safer than during the day, regardless
of size of radar cross-section or other aircraft-specific

The casualty data also show that after the first few days of the war,
the number of night casualties fell off considerably.  Of the 17
identifiable nighttime casualties, all but 3 occurred during the
first 6 days of the war.  There are two plausible, complementary
explanations for this.  First, by day five, the IADS and radar SAMs,
which were unaffected by time of day, had been rendered ineffective
through a combination of actual destruction to radar facilities and
deterrence in turning radars on, achieved through bombing.  Second,
after day three, most low-altitude attacks, and their lower survival
rates, were terminated.  Thus, by the end of the first week, the only
air defense weapon that was not impeded at night--radar SAMs--had
been suppressed and the optically aimed AAA and IR SAMs that were
impeded by night were reduced in effectiveness by the coalition's use
of high-altitude tactics. 

In effect, the data indicate that most Desert Storm aircraft
casualties occurred during the day.  Therefore, it is simply less
likely that any aircraft, including the F-117, which operated only at
night, would have been hit or lost, especially after radar SAMs were
suppressed and low-altitude attacks were discontinued. 

------------------------------------------------------ Appendix II:4.4

Because no F-117s were lost or damaged in Desert Storm, they have
been thought of as uniquely survivable, compared to other aircraft. 
Indeed, the Air Force contended in its September 1991 Desert Storm
white paper that "the F-117 was the only airplane that planners dared
risk over downtown Baghdad" where air defenses are claimed to have
been uniquely dense or severe.\41

\41 USAF, Reaching Globally, Reaching Powerfully:  The United States
Air Force in the Gulf War
(Sept.  1991), p.  56. 

---------------------------------------------------- Appendix II:4.4.1

More radar-guided SAM systems were deployed to the Baghdad area than
any other area in Iraq, and diagrams of SAM coverage confirm that the
greatest concentration of defenses were in that area.  (Table II.10
presents the number and location of Iraqi SAM batteries.)

                              Table II.10
                    Number and Location of Iraqi SAM

                                 SA-   SA-   SA-   SA-
Location                           2     3     6     8  Roland   Total
------------------------------  ----  ----  ----  ----  ------  ======
Mosul/Kirkuk                       1    12     0     1       2      16
H-2/H-3                            1     0     6     0       6      13
Talil/Jalibah                      1     0     0     0       2       3
Basrah                             2     0     8     0       5      15
Baghdad                           10    16     8    15       9      58
Source:  USAF, History of the Air Campaign, p.  254. 

However, it is relevant to note that the defense systems located in
the Baghdad area did not necessarily protect downtown Baghdad at a
higher threat level than the rest of the overall metropolitan area. 
This would be logical, since likely targets for any of Iraq's
adversaries were not only downtown, but were dispersed--along with
radar SAM sites--throughout the Baghdad area.  The distribution of
radar SAMs deployed to the overall Baghdad region is shown in figure
II.7.  These include SA-2, SA-3, SA-6, SA-8, and Roland missiles. 

   Figure II.7:  Radar-Guided SAM
   Locations in the Baghdad Area

   (See figure in printed

   Source:  52nd Fighter Wing
   Desert Storm, A success story ,
   Briefing, GWAPS Files, GWAP,
   vol.  IV, pt.  I:  Weapons,
   Tactics, and Training Report
   (unclassified), p.  12.

   (See figure in printed

The greatest concentrations of radar SAMs were clearly not in the
center of the city but, rather, in its outlying regions.  The lethal
range of these systems was described by Air Force intelligence
experts as extending over the general Baghdad area, as far as 60
miles outside the city. 

Moreover, because the engagement range of the five different types of
SAMs varied, and because they were dispersed throughout the Baghdad
area, it appears unlikely that they somehow converged over the
downtown area to make it the most dangerous locus of all.  The
maximum engagement ranges of the systems varied from 3.5 miles for
the Roland to 27 miles for the SA-2.\42 Only the Vietnam-era vintage
SA-2s would have had sufficient range to cover most of the area shown
in figure II.7 and to converge over the center of the city.\43

For the others, with ranges varying from 3.5 to 13 miles, the
deployment pattern shows that the densest concentrations of
overlapping radar SAM defenses were outside downtown Baghdad. 


With regard to the two other principal antiaircraft defenses, IR SAMS
and AAA, there were clearly more AAA sites in the Baghdad area than
elsewhere in Iraq, but IR SAMs were deployed only to army field
units, mostly in the KTO and not at all in Baghdad.  (See fig. 

However, AAA sites, like radar SAMs, were deployed throughout the
greater metropolitan Baghdad area, not just downtown.  Therefore,
while AAA in the Baghdad region may have been more severe than
elsewhere, it is also the case that it endangered not just the F-117s
but all other coalition aircraft that conducted strikes in the
general metropolitan area. 

   Figure II.8:  AAA Deployment in
   Iraq\ a

   (See figure in printed


\a Does not include IR SAMs and AAA deployed to Iraqi army and
Republican Guard forces in the field. 

Source:  GWAPS, vol.  II, pt.  I:  Operations Report (Secret), p. 

\42 According to USAF intelligence data, the maximum ranges were
SA-2, 27 miles; [DELETED]; Roland, 3.5 miles. 

\43 Also, the SA-6, with the next greatest assessed range of these
systems, is at least 20 years old and [DELETED]. 

---------------------------------------------------- Appendix II:4.4.2

Given the distribution of defenses throughout the Baghdad region, the
survivability of the F-117 is more appropriately compared to that of
other aircraft that were tasked to targets in the region, not just to
those tasked to downtown targets.\44 In this context, we found that
five other types of aircraft made repeated strikes in the Baghdad
region--F-16s, F/A-18s, F-111Fs, F-15Es, and B-52s.  Large packages
of F-16s were explicitly tasked to "downtown" targets in the first
week of the air campaign, but these taskings were stopped after two
F-16s were lost to radar SAMs over the Baghdad area during daytime. 
Available data report no casualties over the Baghdad area, except for
one F/A-18, one GR-1, and the two F-16s cited above.\45

Assertions that the F-117 was uniquely survivable because it alone
was tasked to uniquely severe defenses over downtown Baghdad are
therefore not supported by the data.  F-117s never faced the defenses
that proved to be the most lethal in Desert Storm--daytime AAA and IR
SAMs.  Whereas, the defenses around metropolitan Baghdad were among
the most potent in Iraq, the defenses over downtown were not more
severe than those over the metropolitan area.  Other aircraft were
tasked to equally heavily defended targets.  Moreover, some aircraft
that flew at night also conducted strikes without casualties. 

In sum, the factor most strongly associated with survivability in
Desert Storm appears to have been the combination of flying high and
flying at night--an environment that the F-117s operated in

\44 Other aircraft that were tasked to Baghdad and attacked during
the day would have faced more severe defenses than did the F-117s at
night:  during the day optically aimed AAA would be able to operate
at its most effective level. 

\45 GWAPS and other reports did not specify the locations of all
aircraft casualties.  Therefore, it is possible that some of these
aircraft were damaged or lost over the Baghdad metropolitan area, but
the data available do not specify locations. 

------------------------------------------------------ Appendix II:4.5

Two additional factors are notable about aircraft survivability from
available data. 

---------------------------------------------------- Appendix II:4.5.1

One early tactic in the air war that may have had the effect of
causing some aircraft losses to Iraqi defenses was to send large
numbers of aircraft over a target one after another.  While the first
aircraft over the target frequently encountered no defenses, its bomb
detonations would alert the Iraqis, resulting in AAA and SAMs being
directed against the aircraft that followed.  [DELETED]

---------------------------------------------------- Appendix II:4.5.2

The fact that the optically aimed AAA and IR SAMs remained lethal
throughout the air campaign put a premium on the extent to which
aircraft operating during the day could be made less visible through
camouflage.  A-10 pilots told us that the aircraft's dark green paint
scheme--intended for low-level operations in northern Europe
(including for concealment from aircraft from above)--made them stand
out in the desert against both sand and sky.  Consequently, some A-10
units began to paint their aircraft the same light grey color scheme
of most other Air Force aircraft.  However, the units that repainted
their A-10s were subsequently ordered by Air Force Component, Central
Command (CENTAF) to change them back to dark green. 

A total of 20 A-10s was hit during the war--nearly 25 percent of all
aircraft casualties.  Some A-10 pilots we spoke to believed--and one
participating unit's after-action report stated--that the dark green
paint was unacceptable and may have been responsible for some of the
casualties.  A postwar Air Force study on survivability stated that
the concerns over the A-10's paint scheme were "valid" and
recommended that, in the future, "Paint schemes must be adaptive to
the environment in which the aircraft operate."\46 It is noteworthy
that no A-10s were shot down, or even damaged, at night, when the
dark paint scheme very probably assisted them or, at minimum, did not
make them stand out. 

\46 USAF Air Warfare Center, U.S.  Air Force Surface-to-Air
Engagements During Operation Desert Storm (Secret), Eglin Air Force
Base:  January 1992, p.  12. 

-------------------------------------------------------- Appendix II:5

In this appendix, we addressed questions concerning pre-Desert Storm
claims made for air-to-ground aircraft, munitions, and target sensor
systems versus how they were actually used in Desert Storm.  In
addition, we examined trends in aircraft and munition use, with
particular emphasis on the F-117, and aircraft survivability,
including the factors suggested by Desert Storm data that are most
likely to account for aircraft casualties. 

We first examined the operating environment of Desert Storm to
provide the relevant context.  The coalition faced a well-understood
threat and had considerable lead time to prepare and actually
practice for the eventual conflict.  This provided coalition forces
with an edge that should not be discounted in evaluations of the
outcomes of the Persian Gulf War.  The coalition had 6 months to plan
for the war, deploy the necessary assets to the theater, practice
strikes and deceptions, gather intelligence on targets, and become
highly familiar with the operating environment.  The fact that the
coalition knew which IADS nodes to hit to inflict the most damage,
the most quickly, was critical to its rapid degradation, and to the
achievement of a form of air supremacy--elimination of an integrated,
coordinated air defense.  Without this supremacy, the air campaign
might have proceeded at a much slower pace and perhaps with more
losses.  Further, the United States had the advantage of facing a
highly isolated adversary, essentially unable to be reinforced by
air, sea, or ground.  The unique and often cooperative conditions of
Desert Storm also severely limit the lessons of the war that can be
reasonably applied to potential future contingencies. 

We next compared planned aircraft and munitions use to actual Desert
Storm use, along with patterns of aircraft and munition weight of
effort against sets of strategic targets.  While there were few
notable discrepancies between original aircraft or munitions design
and actual use of either in the conflict, two that are related did
stand out:  the survivability decision to bar munitions deliveries
from below 12,000 feet after day 2 and the corresponding fact that
most unguided munitions tactics, before the war, planned for
low-altitude deliveries.  The switch to medium- to high-altitude
deliveries meant that the accuracy of unguided munitions was greatly
reduced.  This trade-off was feasible in Desert Storm as a way to
reduce attrition--in fact, to almost eliminate it.  But since 95
percent of the bombs and 92 percent of the total tonnage were
unguided, there may have been a severe reduction in the accuracy of
that ordnance. 

In less than half of the strategic target categories, there was a
clear preference for a particular type of air-to-ground platform. 
Preferences were evidenced for F-117s, F-15Es, A-6Es, and F/A-18s
against C\3 , GVC, NBC (F-117), NAV (A-6E and F/A-18), and SCU
(F-15E) targets.  Nonetheless, considering all target categories and
selected platforms, most aircraft were assigned to multiple targets
across multiple target categories. 

The combination of the ban on low-altitude tactics after day two, the
degradation of radar SAMs and the IADS in the early days of the war,
and the fact that a high proportion of strikes were flown at
night--which constituted another form of aircraft sanctuary--almost
certainly was responsible for a coalition aircraft attrition rate
well below what planners expected and below historical precedent in
the Middle East. 

The Desert Storm air campaign was not accomplished by the efforts of
strike aircraft alone.  Aerial refueling tankers, airborne
intelligence-gathering aircraft, reconnaissance aircraft, and strike
support aircraft like F-4Gs, F-15Cs, and EF-111s were vital
ingredients in the successful execution of the air campaign. 

While many factors about the operating environment in Desert Storm
were highly favorable to the coalition's air effort, aircraft
targeting capabilities and precision munitions were put to the test
by some periods of adverse weather as well as adverse conditions like
smoke from oil fires or dust from bombing.  Even mild weather
conditions, including humidity, rendered precision bombing sensors
(such as IR target detection systems and laser target designation
systems) either degraded or unable to work at all.  Moreover, even in
clear weather, pilots sometimes found it difficult to locate or
identify valid targets from medium and high altitudes.  In sum, our
research and analysis found that official DOD descriptions of
aircraft targeting capabilities were overstated based on the Desert
Storm experience. 

Finally, we addressed the role of the F-117 in the Desert Storm air
campaign and examined some of the significant controversies about its
use and contribution.  Contrary to their "Lone Ranger" image, F-117s
certainly required tanking as well as radar jamming support, while
support from air-to-air fighter aircraft is less clear.  The claim
that F-117s--often, but not always--achieved tactical surprise, as
defined by the absence of AAA until bombs made impact, was matched by
the experience of other aircraft.  The gains provided by stealthiness
also required substantial trade-offs in terms of capabilities and
flexibility, including [DELETED].  No F-117s were reported lost or
damaged in Desert Storm, but they operated exclusively at night and
at medium altitudes.  This operational context was clearly less
likely to result in aircraft casualties than low-level attacks or
attacks at any level in daylight.  Moreover, like the F-117s, some
other nonstealth attack aircraft experienced no losses operating in
the high-threat areas of Baghdad and operating at night at medium