[Back] [Index] [Next]

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


Operation Desert Storm:
Evaluation of the Air Campaign

LETTER


B-260509

July 2, 1996

The Honorable David Pryor
Committee on Governmental Affairs
United States Senate

The Honorable John D.  Dingell
Ranking Minority Member
Committee on Commerce
House of Representatives

This report responds to your request that we comprehensively evaluate
the use and effectiveness of the various aircraft, munitions, and
other weapon systems used in the victorious air campaign in Operation
Desert Storm in order to aid the Congress in future procurement
decisions. 

Over 5 years ago, the United States and its coalition allies
successfully forced Iraq out of Kuwait.  The performance of aircraft
and their munitions, cruise missiles, and other air campaign systems
in Desert Storm continues to be relevant today as the basis for
significant procurement and force sizing decisions.  For example, the
Department of Defense (DOD) Report on the Bottom-Up Review (BUR)
explicitly cited the effectiveness of advanced weapons used in Desert
Storm--including laser-guided bombs (LGB) and stealth aircraft--as
shaping the BUR recommendations on weapons procurement.\1


--------------------
\1 Department of Defense, Report on the Bottom-Up Review (Washington,
D.C.:  Oct.  1993), p.  18. 


   BACKGROUND
------------------------------------------------------------ Letter :1

Operation Desert Storm was primarily a sustained 43-day air campaign
by the United States and its allies against Iraq between January 17,
1991, and February 28, 1991.  It was the first large employment of
U.S.  air power since the Vietnam war, and by some measures
(particularly the low number of U.S.  casualties and the short
duration of the campaign), it was perhaps the most successful war
fought by the United States in the 20th century.  The main ground
campaign occupied only the final 100 hours of the war. 

The air campaign involved nearly every type of fixed-wing aircraft in
the U.S.  inventory, flying about 40,000 air-to-ground and 50,000
support sorties.\2 Approximately 1,600 U.S.  combat aircraft were
deployed by the end of the war.  By historical standards, the
intensity of the air campaign was substantial.  The U.S.  bomb
tonnage dropped per day was equivalent to 85 percent of the average
daily bomb tonnage dropped by the United States on Germany and Japan
during the course of World War II. 

Operation Desert Storm provided a valuable opportunity to assess the
performance of U.S.  combat aircraft and munitions systems under
actual combat conditions.  Unlike operational tests or small-scale
hostilities, the air campaign involved a very large number of
conventional systems from all four services used in tandem, which
permits potentially meaningful cross-system comparisons.  The combat
data in this report can be seen as an extension of the performance
data generated by DOD's operational test and evaluation programs that
we have previously reviewed.\3


--------------------
\2 Support sorties comprised missions such as refueling, electronic
jamming, and combat air patrol. 

\3 See Weapons Acquisition:  Low-Rate Initial Production Used to Buy
Weapon Systems Prematurely (GAO/NSIAD-95-18, Nov.  21, 1994); Weapons
Acquisition:  A Rare Opportunity for Lasting Change (GAO/NSIAD-93-15,
Dec.  1992); Weapons Testing:  Quality of DOD Operational Testing and
Reporting (GAO/PEMD-88-32BR, July 26, 1988); Live Fire Testing: 
Evaluating DOD's Programs (GAO/PEMD-87-17, Aug.  17, 1987); and How
Well Do the Military Services Perform Jointly in Combat?  DOD's Joint
Test and Evaluation Program Provides Few Credible Answers
(GAO/PEMD-84-3, Feb.  22, 1984). 


   OBJECTIVES, SCOPE, METHODOLOGY
------------------------------------------------------------ Letter :2

To respond to your questions about the effectiveness of the air
campaign; the performance of individual weapon systems; the accuracy
of contractor claims, particularly in regard to stealth technology
and the F-117; and the relationship between the cost of weapon
systems and their performance and contributions to the success of the
air campaign, we established the following report objectives. 

1.  Determine the use, performance, and effectiveness of individual
weapon systems in pursuit of Desert Storm's objectives and, in
particular, the extent to which the data from the conflict support
the claims that DOD and weapon contractors have made about weapon
system performance. 

2.  Describe the relationship between cost and performance for the
weapon systems employed. 

3.  Identify the degree to which the goals of Desert Storm were
achieved by air power. 

4.  Identify the key factors aiding or inhibiting the effectiveness
of air power. 

5.  Identify the contributions and limitations of advanced
technologies to the accomplishments of the air campaign. 

6.  Determine whether the unique conditions of Desert Storm limit the
lessons learned. 

We compared the performance of nine fixed-wing air-to-ground aircraft
and assessed several major guided and unguided bombs and missiles
used in the war, including Tomahawk land attack (cruise) missiles
(TLAM), laser-guided bombs (LGB), Maverick missiles, and unitary
unguided bombs.\4 The primary focus of our analysis was on the use of
these weapon systems in missions against targets that war planners
had identified as strategic.\5

Historically, studies of air power have articulated differing points
of view on the relative merits of focusing air attacks on targets
deemed to be strategic (such as government leadership, military
industry, and electrical generation) and focusing them on tactical
targets (such as frontline armor and artillery).  These contending
points of view have been debated in many official and unofficial
sources.\6 In this study, we did not directly address this debate
because data and other limitations (discussed below) did not permit a
rigorous analysis of whether attacks against strategic targets
contributed more to the success of Desert Storm than attacks against
tactical targets. 

A primary goal of our work was to cross-validate the best available
data on aircraft and weapon system performance, both qualitative and
quantitative, to test for consistency, accuracy, and reliability.  We
collected and analyzed data from a broad range of sources, including
the major DOD databases that document the strike histories of the war
and cumulative damage to targets; numerous after-action and
lessons-learned reports from military units that participated in the
war; intelligence reports; analyses performed by DOD contractors;
historical accounts of the war from the media and other published
literature; and interviews with participants, including more than 100
Desert Storm pilots and key individuals in the planning and execution
of the war.\7 And after we collected and analyzed the air campaign
information, we interviewed DOD, Joint Chiefs of Staff (JCS), and
service representatives and reviewed plans for the acquisition and
use of weapon systems in future campaigns to observe how the lessons
learned from Desert Storm have been applied. 

To compare the nature and magnitude of the power that Operation
Desert Storm employed against strategic targets to the nature of
outcomes, we analyzed two databases--the "Missions" database
generated by the Air Force's Gulf War Air Power Survey (GWAPS)
research group to assess inputs and the Defense Intelligence Agency's
(DIA) phase III battle damage assessment (BDA) reports to assess
outcomes.  While this methodology has limitations, no other study of
Desert Storm has produced the target-specific, input-outcome data
that can be derived by merging these databases. 

The data we analyzed in this report constitute the best information
collected during the war.\8 We focused our analyses on data available
to commanders during the war--information they used to execute the
air campaign.  These data also provided the basis for many of the
postwar DOD and manufacturer assessments of aircraft and weapon
system performance during Desert Storm.\9


--------------------
\4 The aircraft included the A-6E, A-10, B-52, F-16, F-15E, F/A-18,
F-111F, and F-117 from the U.S.  air forces, as well as the British
GR-1.  The AV-8B, A-7, and B-1B were not included.  Both the AV-8B
and the A-7 were excluded because of their relatively few strikes
against strategic targets.  The B-1B did not participate in the
campaign because munitions limitations, engine problems, inadequate
crew training, and electronic warfare deficiencies severely hampered
its conventional capabilities. 

\5 Campaign planners categorized all strategic targets into 1 of 12
target sets:  command, control, and communication (C\3 ); electrical
(ELE); government centers or leadership (GVC); lines of communication
(LOC); military industrial base (MIB); naval (NAV); nuclear,
biological, and chemical (NBC); offensive counterair (OCA); oil
refining, storage, and distribution (OIL); Republican Guard (RG) or
ground order of battle (GOB); surface-to-air missile (SAM); and Scud
missile (SCU). 

\6 Examples include Edward C.  Mann, III, Thunder and Lightning
(Maxwell Air Force Base, Ala.:  Air University Press, Apr.  1995);
John A.  Warden, III, The Air Campaign (Washington, D.C.: 
Pergamon-Brassey's, 1989); and Richard T.  Reynolds, Heart of the
Storm:  The Genesis of the Air Campaign Against Iraq (Maxwell Air
Force Base, Ala.:  Air University Press, Apr.  1995). 

\7 We interviewed pilots representing each type of aircraft
evaluated, with the exception of British Tornados.  The British
government denied our requests to interview British pilots who had
flown in Desert Storm. 

\8 We also sought data and analyses collected and conducted after the
war.  We used these data to check the reliability and validity of
information collected earlier. 

\9 Constraints in the reliability and completeness of some important
portions of the data imposed limitations on our analysis of the air
campaign.  For example, relating specific types of aircraft or
munitions to target outcomes was problematic because BDA reports
provided a comprehensive compilation of damage on strategic targets
at given times during the campaign--not necessarily after each strike
against the targets.  Therefore, we balanced data limitations, to the
extent possible, through qualitative analyses of systems, based on
the diverse sources cited above.  For example, we compared claims
made for system performance and contributions to what was supportable
given all the available data, both quantitative and qualitative. 
(See app.  I for additional information on the study methodology and
the strengths and limitations of the data.)


   DATA LIMITATIONS
------------------------------------------------------------ Letter :3

The best available data did not permit us to either (1) make a
comprehensive system-by-system quantitative comparison of aircraft
and weapon effectiveness or (2) validate some of the key performance
claims for certain weapon systems from the war.  However, we were
able to compare aircraft and munition performance in Desert Storm
using a combination of quantitative and qualitative data.  There are
major limitations in the available data pertaining to the effects of
aircraft and munitions on targets.  At the same time, DOD
successfully collected a large amount of data across a wide range of
issues, including weapon use, aircraft survivability, sortie rates,
and support needs.  With the caveats stated above, these data
permitted us to analyze aircraft and weapon system performance,
performance claims, and the effectiveness of air power.\10


--------------------
\10 See appendix I for an expanded discussion of our methodology. 
Appendixes II through XI present the analyses in support of our
findings.  A description of aircraft and munition use is presented in
appendix II.  Appendix III discusses aircraft and munition
performance and effectiveness.  Cost and performance of aircraft and
munitions are analyzed in appendix IV.  The development of air
campaign objectives and the Iraqi air defense system are described in
appendixes V and VI, respectively.  Appendix VII compares the design
mission of aircraft with their actual use, while the weight and types
of effort expended are summarized in appendix VIII.  Supplementary
information on target sensor technologies and combat support
platforms are presented in appendixes IX and X.  Finally, an
examination of the employment of the F-16 and F-117 against the
Baghdad Nuclear Research Facility is presented in appendix XI. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :4

Air power clearly achieved many of Desert Storm's objectives but fell
short of fully achieving others.\11 The available quantitative and
qualitative data indicate that air power damage to several major
target sets was more limited than DOD's title V report to the
Congress stated.\12 These data show clear success against the oil and
electrical target categories but less success against Iraqi air
defense; command, control, and communications, and lines of
communication.  Success against nuclear-related, mobile Scud, and RG
targets was the least measurable. 

The lessons that can be learned from Desert Storm are limited because
of the unique conditions, the strike tactics employed by the
coalition, the limited Iraqi response, and limited data on weapon
system effectiveness.  The terrain and climate were generally
conducive to air strikes, and the coalition had nearly 6 months to
deploy, train, and prepare.  The strong likelihood of campaign
success enabled U.S.  commanders to favor strike tactics that
maximized aircraft and pilot survivability rather than weapon system
effectiveness.  In addition, the Iraqis employed few, if any,
electronic countermeasures and presented almost no air-to-air
opposition.  As a result, Desert Storm did not consistently or
rigorously test all the performance parameters of aircraft and weapon
systems used in the air campaign.  Moreover, as we noted above, data
are not available to fully assess the relative or absolute
effectiveness of aircraft and weapon systems in the war.  This
combination of factors limits the lessons of the war that can be
reasonably applied to future contingencies. 

Many of DOD's and manufacturers' postwar claims about weapon system
performance--particularly the F-117, TLAM, and laser-guided
bombs--were overstated, misleading, inconsistent with the best
available data, or unverifiable. 

Aircraft and pilot losses were historically low, partly owing to the
use of medium- to high-altitude munition delivery tactics that
nonetheless both reduced the accuracy of guided and unguided
munitions and hindered target identification and acquisition, because
of clouds, dust, smoke, and high humidity.  Air power was inhibited
by the limited ability of aircraft sensors to identify and acquire
targets, the failure to gather intelligence on critical targets, and
the inability to collect and disseminate BDA in a timely manner. 
Similarly, the contributions of guided weaponry incorporating
advanced technologies and their delivery platforms were limited
because the cooperative operating conditions they require were not
consistently encountered. 

DOD did not prominently emphasize a variety of systems as factors in
the success of the air campaign.  The important contributions of
stealth and laser-guided bombs were emphasized as was the need for
more and better BDA; less attention was paid to the significant
contributions of less-sophisticated systems and the performance of
critical tasks such as the identification and acquisition of targets. 
For example, more than is generally understood, the air campaign was
aided by relatively older and less technologically advanced weapon
systems and combat support aircraft, such as unguided bombs, the
B-52, the A-10, refueling tankers, and electronic jammer aircraft. 
There was no apparent link between the cost of aircraft and
munitions, whether high or low, and their performance in Desert
Storm. 

After our analysis of the air campaign, we performed a review of the
actions taken by DOD to address the lessons learned from our
findings.  While we found that several lessons were being addressed
by DOD, we also found that others have not been.  The lessons that
have not been fully or appropriately addressed are the subject of
three recommendations at the conclusion of this letter. 


--------------------
\11 The initial objectives of the strategic air campaign were to (1)
disrupt the Iraqi leadership and command and control; (2) achieve air
supremacy; (3) cut supply lines; (4) destroy Iraq's nuclear,
biological, and chemical capability; and (5) destroy the Republican
Guard.  Destroying Scud missiles and mobile launchers became a
priority early in the air campaign. 

\12 Department of Defense, Conduct of the Persian Gulf War, Final
Report to Congress Pursuant to
Title V of the Persian Gulf Conflict Supplemental Authorization and
Personnel Benefits Act of 1991
(P.L.  102-25), April 1992. 


   PRINCIPAL FINDINGS
------------------------------------------------------------ Letter :5


      USE, PERFORMANCE, AND
      EFFECTIVENESS OF AIRCRAFT
      AND WEAPON SYSTEMS
---------------------------------------------------------- Letter :5.1


         AIRCRAFT AND WEAPON
         SYSTEMS USED AS DESIGNED
-------------------------------------------------------- Letter :5.1.1

In general, the actual use of aircraft and weapon systems in the
conflict was consistent with their stated prewar capabilities.  (App. 
II compares in detail the combat mission categories attributed to
each aircraft before Desert Storm and those actually performed during
the campaign.) Most targets were attacked by several types of
aircraft or weapon systems.  However, from strike data and pilot
interviews, we did find that certain aircraft were somewhat preferred
in certain target categories.  The F-117 was the preferred platform
against fixed, often high-value C\3 , leadership, and NBC targets;
against naval targets, the A-6E and F/A-18 were preferred; and
against fixed Scud missile targets, the F-15E.  (The distribution of
strikes by each type of aircraft across each of the strategic target
categories is discussed in app.  II.)

Support aircraft, including refueling tankers, airborne
intelligence-gathering aircraft, reconnaissance aircraft, and strike
support aircraft like the F-4G, F-15C, EF-111, and EA-6B flew more
than 50,000 sorties and were instrumental in the successful execution
of the air campaign.  Each type of strike aircraft, conventional and
stealthy, received support--such as jamming and refueling--although
not necessarily on each mission.  (See app.  II for a discussion of
the support provided to both conventional and stealth aircraft.)


         AIRCRAFT SURVIVABILITY
         ENHANCED BY TACTICS
-------------------------------------------------------- Letter :5.1.2

The aircraft casualty rate (that is, aircraft DOD identified as lost
to Iraqi action or damaged in combat) for the aircraft we reviewed
was 1.7 aircraft per 1,000 strikes.  This rate was very low compared
to planners' expectations and historic experience.  The combination
in the first week of the war of a ban on low-level deliveries for
most aircraft and a successful effort to suppress enemy air defenses
(SEAD) that greatly degraded radar surface-to-air (SAM) missiles and
the Iraqi integrated air defense system (IADS) resulted in a
reduction in the average number of aircraft casualties per day from
6.2 during the first 5 days to about 1.5 for the remaining
38 days of the campaign.  If the aircraft combat casualty rate for
the first
5 days had continued throughout the war, a total of about 267
coalition aircraft would have been casualties.  Avoiding low
altitudes, 48 aircraft were actually damaged in combat during the
entire war, and an additional 38 were combat losses. 

The attrition rate (including both loss and damage) of all combat
aircraft was especially low when they flew at medium and high
altitudes and at night.  For example, only one-third of the Air Force
casualties occurred above 12,000 feet, and only one-quarter of the
coalition aircraft casualties occurred at night.  The attrition rate
at low altitudes was notably higher because of the continuing
presence of antiaircraft artillery (AAA) and portable infrared (IR)
SAMs--systems that are also generally less effective at night. 
Nonetheless, AAA and IR SAMs, perceived before the campaign to be
lesser threats than radar-guided SAMs, were responsible for four
times more casualties than radar SAMs.  (See app.  II for additional
information and analysis on aircraft losses and damage.)

One of the stated advantages of stealth technology is that it
enhances survivability, and in Desert Storm, the stealthy F-117 was
the only aircraft type to incur neither losses nor damage.  However,
these aircraft recorded fewer sorties than any other air-to-ground
platform and flew exclusively at night and at medium altitudes--an
operating environment in which the fewest casualties occurred among
all types of aircraft.\13 Moreover, given the overall casualty rate
of 1.7 per 1,000 strikes, the most probable number of losses for any
aircraft, stealthy or conventional, flying the same number of
missions as the F-117 would have been zero.  (See app.  II for more
information on the tactics and support used by F-117s to minimize
their exposure to air defense threats.)


--------------------
\13 For example, nonstealthy aircraft, such as the F-111F and F-16,
also suffered no losses when operating at night, and the A-10s
experienced neither damage nor losses at night.  Each of these three
aircraft types flew at least as many night strikes as the F-117. 


         GUIDED AND UNGUIDED
         MUNITIONS REVEALED
         STRENGTHS AND WEAKNESSES
-------------------------------------------------------- Letter :5.1.3

While higher altitude deliveries clearly reduced aircraft casualties,
they also caused target location and identification problems for
guided munitions and exposed unguided bombs to uncontrollable factors
such as wind.  Medium- and high-altitude tactics also increased the
exposure of aircraft to clouds, haze, smoke, and high humidity,
thereby impeding IR and electro-optical (EO) sensors and laser
designators for LGBs.  These higher altitude tactics also reduced
target sensor resolution and the ability of pilots to discern the
precise nature of some of the targets they were attacking.  While
pilots and planners reported that unguided bombs were substantially
less accurate and target discrimination problems were sometimes
severe, these unguided bombs were employed with radar against area
targets in poor weather. 

Our interviews with pilots also revealed a mix of concerns about
survivability with guided and unguided munitions.  Pilots pointed out
that in some circumstances, guided munitions permitted the aircraft
to "stand off" at relatively long distances from targets and their
defenses, which was not possible with unguided munitions, while
retaining accuracy.  [DELETED] (See apps.  II and IV for more pilot
views on the use of guided and unguided munitions.)

Guided bombs were the weapon of choice against small, point targets,
such as reinforced bunkers, hardened aircraft shelters, and armored
vehicles.  However, from high altitude, unguided bombs were the
weapon of choice against area targets, such as ammunition storage
facilities and ground troop emplacements.  In addition, pilots,
especially of the F-16, remarked to us that they believed their
high-altitude unguided bomb deliveries were ineffective against point
targets such as tanks. 

Over the course of the campaign, the overall ratio of
guided-to-unguided munitions delivered (1 to 19) did not
significantly change from week to week.  This and other data--such as
interviews with campaign planners and pilots--indicate that there was
no discovery of a systematic failure of either type of munition or
any broad effort to change from one type of munition to another. 
(Patterns of munition use are discussed in app.  II.)


         AIRCRAFT AND MUNITION
         EFFECTIVENESS MEASURES
         DEVELOPED
-------------------------------------------------------- Letter :5.1.4

Despite data limitations in some instances, sufficient data were
generated to permit a limited analysis of the relative effectiveness
of aircraft and munitions.  We developed a surrogate effectiveness
measure by calculating the ratio of fully successful (FS) to not
fully successful (NFS) target outcomes for the set of strategic
targets attacked by each type of weapon system.\14 By comparing these
ratios, we found that effectiveness varied by type of aircraft and by
type of target category attacked.  For example, the F-111F
participated in a higher ratio of FS versus NFS (3.2:1) than any
other aircraft type.  The F-117 and the F-16 performed next best and
at about the same ratio (1.4:1 and 1.5:1, respectively), and the
F-15E and the A-6E both participated in about the same number of
successfully attacked targets as not fully successfully attacked
(1.0:1 and 1.1:1 respectively).\15 Only the B-52 and the F/A-18
participated in more NFS target outcomes than FS (with ratios of
0.7:1 and 0.8:1, respectively).  Data were not available for the
A-10. 

The effectiveness of aircraft and munitions in aggregate varied among
the strategic target sets.\16 While the attainment of strategic
objectives is determined by more than the achievement of individual
target objectives, the compilation of individual target objectives
achieved was one tool used by commanders during the war to direct the
campaign.  Among strategic targets for which BDA were available, the
percent of targets where objectives were successfully met ranged from
a high of 76 percent among (known) nuclear, biological, and chemical
(NBC) targets to a low of 25 percent among fixed Scud-related
strategic targets.\17

No consistent pattern indicated that the key to success in target
outcomes was the use of either guided or unguided munitions.  On
average, targets where objectives were successfully achieved received
more guided and fewer unguided munitions than targets where
objectives were not determined to have been fully achieved.  In
comparing the use of guided munitions to unguided munitions, on
average, approximately 11 tons of guided munitions were delivered
against FS targets and over 9 tons were released against NFS targets. 
Fewer unguided munitions were used against FS targets (44 tons) than
NFS (54 tons).  However, neither pattern held across all target
categories.  In four target categories, NFS targets received more
tons of guided munitions than successful ones, and in six categories,
successful targets received more unguided munitions than the NFS
ones.  (Our complete analysis of air campaign inputs [that is,
numbers and types of aircraft and munitions] and target outcomes
[that is, successfully or not fully successfully met target
objectives] is presented in app.  III.)


--------------------
\14 Using intelligence gathered during the war from multiple sources,
DIA conducted BDA on 357 of the 862 strategic targets in the GWAPS
Missions database.  We categorized the outcomes for these 357
strategic targets as being either fully successful or not fully
successful.  We classified a target outcome as FS if the last BDA
report on that target stated that the target objective had been met
and a restrike was not necessary.  We classified all other target
outcomes as NFS.  DIA produced BDA during the war at the request of
U.S.  Central Command (CENTCOM).  Thus, although the
representativeness of the targets assessed by DIA is unknowable,
these 357 do represent the set of targets of greatest interest to the
commanders in the theater.  (See app.  I for a more detailed
discussion of our BDA classification methodology.)

\15 Although the F-111F participated in the highest ratio of FS to
NFS target outcomes, the F-117 participated in the highest number of
successful outcomes.  The F-117 participated in 122 FS outcomes (as
well as 87 NFS); the next 2 aircraft with the highest participation
in successful outcomes were the F-16, with 67 (and 45 NFS), and the
F-111F, with 41 (and 13 NFS). 

\16 The number of targets in each strategic target set where the
target objectives had been successfully met was used as a measure of
the effectiveness of aircraft and munitions in the aggregate. 
Whether a target objective had been met was determined from the final
DIA phase III BDA report written on a target during the campaign. 

\17 Less than 15 percent of the nuclear-related facilities were
identified before the end of the air campaign. 


         SOME DOD AND CONTRACTOR
         CLAIMS OVERSTATED
-------------------------------------------------------- Letter :5.1.5

As requested, we analyzed numerous Desert Storm performance claims
and found from the available data that DOD, individual military
services, and manufacturers apparently overstated the Desert Storm
performance of certain aircraft and weapon systems that used advanced
technologies.  We found justification in several instances for the
congressional concern that some contractor claims may have been
overstated.  For example, some key claims concerning the F-117, the
TLAM, and LGBs, among other advanced systems, were either misleading,
inconsistent with available data, or unverifiable because of the
absence of data. 

F-117s.  DOD's title V report stated that 80 percent of the bombs
dropped by F-117s hit their target--an accuracy rate characterized by
its primary contractor, Lockheed, as "unprecedented." However, in
Desert Storm, (1) approximately one-third of the reported F-117 hits
either lacked corroborating support or were in conflict with other
available data; (2) the probability of bomb release for a scheduled
F-117 mission was only 75 percent; and (3) for these reasons and
because of uncertainty in the data, the probability of a target's
being hit from a planned F-117 strike in Desert Storm ranged between
41 and 60 percent.\18 Similarly, (1) F-117s were not the only
aircraft tasked to targets in and around Baghdad where the defenses
were characterized as especially intense, (2) F-117s were neither as
effective on the first night of the war as claimed nor solely
responsible for the collapse of the Iraqi IADS in the initial hours
of the campaign, (3) F-117s did not achieve surprise every night of
the campaign, and (4) F-117s occasionally benefited from jammer
support aircraft.  (Analyses of F-117 bomb hit data are presented in
app.  III; the ability of F-117 stealth fighters to achieve tactical
surprise is discussed in app.  II.)

TLAMs.  While TLAMs possess an important characteristic distinct from
any aircraft in that they risk no pilot in attacking a target, they
can be compared to aircraft on measures such as accuracy and
survivability.  Their accuracy was less than has been implied.  The
DOD title V report stated that the "launching system success rate was
98 percent." However, this claim is misleading because it implies
accuracy that was not realized in Desert Storm.  Data compiled by the
Center for Naval Analyses (CNA) and DIA in a joint study revealed
that only [DELETED] percent of the TLAMs arrived over their intended
target area, and only [DELETED] percent actually hit or damaged the
intended aimpoint.\19

From [DELETED] TLAMs were apparently lost to defenses or to system
navigation flaws.  Thus, the TLAMs experienced an en route loss rate
as high as [DELETED] percent.\20 (See app.  III for a more detailed
analysis of TLAM performance.)

LGBs.  The manufacturer of the most advanced LGB guidance system
(Paveway III) claimed that it has a "one target, one bomb"
capability.  DOD officials adopted the phraseology to demonstrate the
value of advanced technology in Desert Storm.  We sampled Paveway III
LGB targets and found that the "one target, one bomb" claim could not
be validated, as no fewer than two LGBs were dropped on each target. 
Six or more were dropped on 20 percent of the targets, eight or more
were dropped on 15 percent of the targets, and the overall average
dropped was four LGBs per target.  And larger numbers of Paveway III
and other LGB types were dropped on other targets.  Moreover, as
noted earlier, an average of approximately 11 tons of guided
munitions--most of them LGBs--were used against targets that DIA's
phase III BDA messages showed were successfully attacked.  This
notwithstanding, the number of LGBs required for point targets was
clearly less than the number of unguided munitions needed in this and
previous wars, especially from medium and high altitudes.  (See app. 
III for our analysis of the "one target, one bomb" claim.)

Table 1 shows some of the discrepancies between the claims and
characterizations of manufacturers to the Congress and the public
about the actual and expected performance of weapon systems in combat
and what the data from Desert Storm support.  (App.  III contains
additional examples of discrepancies between manufacturers' claims
and our assessment of weapon system performance in Desert Storm.)



                                     Table 1
                     
                     Manufacturers' Statements About Product
                       Performance Compared to Our Findings

Manufacturer      Their statement               Our finding
----------------  ----------------------------  --------------------------------
General Dynamics  "No matter what the [F-16]    The F-16's delivery of guided
                  mission, air-to-air, air-     munitions, such as Maverick, was
                  to-ground. No matter what     impaired and sometimes made
                  the weather, day or night."   impossible by clouds, haze,
                                                humidity, smoke, and dust. Only
                                                less accurate unguided munitions
                                                could be employed in adverse
                                                weather using radar.

Grumman           "A-6s . . . [were]            The A-6E FLIR's ability to
                  detecting, identifying,       detect and identify targets was
                  tracking, and destroying      limited by clouds, haze,
                  targets in any weather, day   humidity, smoke, and dust; the
                  or night."                    laser designator's ability to
                                                track targets was similarly
                                                limited.\a Only less accurate
                                                unguided munitions could be
                                                employed in adverse weather
                                                using radar.

Lockheed          "During the first night, 30   On the first night, 21 of the 37
                  F-117s struck 37 high-value   targets to which F-117s were
                  targets, inflicting damage    tasked were reported hit; of
                  that collapsed Saddam         these, the F-117s missed
                  Hussein's air defense system  40 percent of their air defense
                  and all but eliminated        targets. BDA on 11 of the F-117
                  Iraq's ability to wage        strategic air defense targets
                  coordinated war."             confirmed only 2 complete kills.
                                                Numerous aircraft, other than
                                                the F-117, were involved in
                                                suppressing the Iraqi IADS,
                                                which did not show a marked
                                                falloff in aircraft kills until
                                                day five.

Martin Marietta   Aircraft with LANTIRN\ can    The LANTIRN can be employed
                  "locate and attack targets    below clouds and weather;
                  at night and under other      however, its ability to find and
                  conditions of poor            designate targets through
                  visibility using low-level,   clouds, haze, smoke, dust, and
                  high speed tactics."\b        humidity ranged from limited to
                                                no capability at all.

McDonnell         TLAMs "can be launched . . .  The TLAM's weather limitation
Douglas           in any weather."              occurs not so much at the launch
                                                point but in the target area
                                                where the optical [DELETED].

Northrop          The ALQ-135 "proved itself    [DELETED]
                  by jamming enemy threat
                  radars"; and was able "to
                  function in virtually any
                  hostile environment."

Texas             "TI Paveway III: one target,  Of a selected sample of 20
Instruments       one bomb."                    targets attacked by F-117s and
                                                F-111Fs with GBU-24s and GBU-
                                                27s, no single aimpoint was
                                                struck by only 1 LGB--the
                                                average was 4, the maximum 10.
--------------------------------------------------------------------------------
\a Forward-looking infrared (FLIR). 

\b Low-altitude navigation and targeting infrared for night
(LANTIRN). 


--------------------
\18 A planned strike is the tasking of one or more bombs against a
specific aimpoint or target on a scheduled F-117 mission as recorded
in the official 37th Tactical Fighter Wing (TFW) Desert Storm
database. 

\19 This analysis addresses TLAM C and D-I models only; data on the
D-II model were excluded because of classification issues. 

\20 Beyond TLAM's [DELETED]-percent miss rate against intended
targets, it demonstrated additional problems.  The relatively flat,
featureless, desert terrain in the theater made it difficult for the
Defense Mapping Agency (DMA) to produce usable Terrain Contour
Matching (TERCOM) ingress routes, and TLAM demonstrated limitations
in range, mission planning, lethality, and effectiveness against hard
targets and targets capable of mobility.  Since the war, the Navy has
developed a Block III variant of the TLAM.  Its improvements include
the use of Global Positioning System (GPS) in TLAM's guidance system. 
With GPS, TLAM route planning is not constrained by terrain features,
and mission planning time is reduced.  However, some experts have
expressed the concern that GPS guidance may be vulnerable to jamming. 
Thus, until system testing and possible modifications demonstrate
TLAM
Block III resistance to electronic countermeasures, it is possible
that the solution to the TERCOM limitations--GPS--may lead to a new
potential vulnerability--jamming.  Moreover, the Block III variant
continues to use the optical Digital Scene Matching Area Correlator
(DSMAC), which has various limitations.  [DELETED]


         DATA INADEQUATE FOR
         COMPREHENSIVE AIRCRAFT
         AND WEAPON SYSTEM
         COMPARISONS OR VALIDATION
         OF SOME CLAIMS
-------------------------------------------------------- Letter :5.1.6

The data compiled on campaign inputs (that is, use of weapon systems)
and outcomes (that is, battle damage assessments) did not permit a
comprehensive effectiveness comparison of aircraft and weapon
systems.  The most detailed Desert Storm strike history summary is
less than complete, does not provide outcome information
consistently, and does not provide strike effectiveness information. 
For example, because data on a large number of A-10 strike events
were unclear or contradictory, we found it impossible to reliably
analyze and include A-10 strike data.\21 In addition, the most
comprehensive BDA database is less than complete, is constrained by
technological limitations associated with imagery intelligence, and
in most cases did not benefit from ground verifications or damage
updates after the war.  Because multiple aircraft of different types
delivered multiple bombs of different types, often on the same
aimpoint, and because damage was often not assessed until after
multiple strikes, it is not possible to determine for most targets
what effects, if any, can be attributed to a particular aircraft or
particular munition.  Moreover, DIA conducted BDA on only 357 of the
862 strategic targets in our analysis for which strike data were
available.  Therefore, many questions on the effectiveness of
aircraft and missile strikes could not be answered nor could some
effectiveness claims.  (For additional information on data
limitations, see apps.  I and III.)


--------------------
\21 This was significant for two reasons.  First, the data that are
available on the A-10 imply that it may have performed even more than
the large number of sorties currently attributed to it.  Second,
because the A-10 was a major participant in the air war and because
it performed at relatively high levels on measures such as sortie
rate and payload, it would have been useful to be able to compare its
success rate, particularly as a low-cost aircraft, against targets to
the other aircraft under review. 


      RELATIONSHIP BETWEEN COST
      AND PERFORMANCE
---------------------------------------------------------- Letter :5.2

Data limitations did not permit a systematic comparison of weapon
system cost and performance; where data were available, our analysis
results either were ambiguous or revealed no consistent trends. 


         PERFORMANCE OF HIGH-COST
         COMPARED TO LOW-COST
         AIRCRAFT
-------------------------------------------------------- Letter :5.2.1

The cost of aircraft was not consistently associated with performance
for several measures such as effectiveness, adverse weather
capability, sortie rate, payload, and survivability.  Survivability
was consistently high for all types of aircraft and therefore
indistinguishable for high- and low-cost aircraft.\22 The high-cost
F-117 stealth fighter and the low-cost A-10 both experienced
100-percent survivability when operating at night.  Although the data
on some measures were ambiguous (such as survivability and
effectiveness), differences in performance or capabilities between
high- and low-cost aircraft were evident for some measures. 

Depending on the measure one uses, aircraft types with different
costs can be characterized as more, less, or equally capable.  For
example, in Desert Storm, average sortie rates and payloads for
different aircraft showed an inverse relationship between cost and
performance.  Moreover, during the campaign, high- and low-cost
aircraft were often employed against the same targets.  Nearly 51
percent of the strategic targets attacked by the stealthy F-117s were
also attacked by less costly, conventional aircraft--such as the
F-16, F-15E, and F/A-18.  The incompleteness of A-10 strike data
prevents our identifying the extent, if any, to which A-10 and F-117
target taskings overlapped.  However, according to GWAPS, both
aircraft performed over 40 strikes in the C\3 , offensive counter
(OCA), SAM, and Scud missile (SCU) strategic target categories.  In
regard to other aircraft, the available strike data reveal that the
F-117 and the F-16 were tasked to 78 common targets, the F-117 and
the F/A-18C/D to 62, and the F-117 and the F-15E to 49. 

Advocates of the F-117 can argue, based on its performance in Desert
Storm, that it alone combined the advantages of stealth and LGBs,
penetrated the most concentrated enemy defenses at will, permitted
confidence in achieving desired bombing results, and had perfect
survivability.  Advocates of the A-10 can, for example, argue that
it, unlike the F-117, operated both day and night; attacked both
fixed and mobile targets employing both guided and unguided bombs;
and like the F-117, suffered no casualties when operating at night
and at medium altitude.  Similarly, other aircraft also performed
missions the F-117 was unable to and were used successfully--and
without losses--against similar types of strategic targets.  Each
aircraft of the various types has both strengths and limitations;
each aircraft can do things the other cannot.  Therefore, despite a
sharp contrast in program unit costs, we find it inappropriate, given
their use, performance, and effectiveness demonstrated in Desert
Storm, to rate one more generally "capable" than the other. 

We also found no consistent relationship between the program unit
cost of aircraft and their relative effectiveness against strategic
targets, as measured by the ratio of FS to NFS target outcomes for
the set of strategic targets that each type of aircraft attacked. 
The high-cost F-111F participated in proportionately more successful
target outcomes than any other aircraft type, but the low-cost F-16
participated in a higher proportion of successful target outcomes
than either the F-117 or the F-15E, both much higher cost aircraft. 
However, the F-117 and the F-111F, two high-cost, LGB-capable
aircraft, ranked first and third in participation against successful
targets.\23 (The complete analysis of the performance of low- and
high-cost aircraft is presented in app.  IV.)


--------------------
\22 Survivability depends on numerous factors, including assistance
from support aircraft, quantity and quality of air defenses, size of
strike package, altitude, and tactics.  In Desert Storm, neither cost
nor stealth technology was found to be a determinant of
survivability. 

\23 Participation by each type of air-to-ground aircraft against
targets assessed as FS targets was as follows:  F-117 = 122; F-16 =
67, F-111F = 41, A-6E = 37, F/A-18 = 36, F-15E = 28, B-52 = 25, and
GR-1 = 21.  No data were available for the A-10.  TLAM participated
against 18 targets assessed as FS.  Participation against FS targets
by type of aircraft is a function of two factors--the breadth of
targets tasked to each type of aircraft (see app.  III) and their
FS:NFS ratio as presented previously. 


         GUIDED MUNITIONS COMPARED
         TO UNGUIDED MUNITIONS
-------------------------------------------------------- Letter :5.2.2

In Desert Storm, 92 percent of the munitions expended were unguided. 
On the assumption that this tonnage contributed to the successful
outcome of the entire campaign--at a minimum by permitting nearly
continuous attacks against both ground force and strategic targets
for 38 days--it is evident that the same campaign accomplishments
would have been difficult or impossible with aircraft dropping
comparatively small numbers of precision-guided munitions (PGM). 

Although only 8 percent of the munitions used against planned targets
were guided, they represented approximately 84 percent of the total
cost of munitions.  The difference in cost between various types of
guided and unguided munitions was quite substantial:  the unguided
unitary bombs used in the air campaign cost, on average, $649 each,
while the average LGB cost more than $30,000 each--a ratio of
1:47.\24 IR Maverick missiles cost about $102,000 each--a cost ratio
to the unguided bombs of 1:157. 

Although cost ratios between guided and unguided weapon systems used
in Desert Storm can be readily calculated, data on the relative
accuracy or effectiveness of the systems in Desert Storm are limited
and often ambiguous.  For example, guided and unguided munitions were
often used against the same targets.  Therefore, given shortfalls in
BDA, a precise probability of kill for munitions could not be
determined in most instances.  However, CNA found a small number of
bridges where conditions and data enabled an assessment of
effectiveness.  These bridges had been attacked with either guided or
unguided bombs, and BDA had been performed in time to distinguish
which type of munitions were successful.  While the sample is small
and cannot be generalized, these data show that (1) substantially
more unguided bombs than either LGBs or Walleyes were required to
successfully destroy a bridge and (2) the cost of the guided
munitions used was substantially higher.\25 (See app.  IV.)

Cost appears to have been a factor in the selection of munitions by
Desert Storm campaign commanders.  For example, some pilots we
interviewed were instructed to use LGBs and Mavericks only against
high-value targets such as tanks, armored personnel carriers, and
artillery (rather than trucks or other GOB targets).  If they could
not hit these targets, they were not able to use these munitions. 
They could, however, drop unguided bombs on other targets before
returning to base.  Similarly, the employment of TLAMs was terminated
after February 1.  GWAPS reported that Gen.  H.  Norman Schwarzkopf,
commander in chief of U.S.  Central Command, approved no additional
TLAM strikes because either (1) television coverage of daylight
strikes in downtown Baghdad proved unacceptable in Washington or (2)
their use was deemed too expensive given the TLAM's relatively small
warhead and high cost.  Thus, this high-cost munition was not used
during the latter two-thirds of the war. 

Increasing the proportion of the U.S.  weapons inventory comprised of
high-cost munitions has potential implications for the future
effectiveness and employment of air power.  First, for a given level
of resources, much higher costs limit the number of weapons that can
be procured.  With fewer weapons, the priority attached to the
survival and successful employment of each high-cost bomb is likely
to be high, as demonstrated in Desert Storm.  Second, Desert Storm
revealed that a focus on increasing aircraft and pilot survivability
may have reduced mission effectiveness, thereby increasing the number
of munitions required to destroy or damage a target.  Third, Desert
Storm showed that commanders were less willing to permit the
widespread use of very expensive munitions; the value of the target
had to be sufficient to justify the cost of a guided weapon. 

Thus, an increasing dependence on high-cost weaponry can lead to
three types of concerns:  limitations in the availability and use of
high-cost systems, the need to increase the munition expenditure rate
per target to compensate for lessened effectiveness when emphasizing
survivability, and a diminished ability to attack large numbers of
targets (such as lower priority GOB).\26 (See app.  IV for further
discussion of the performance of high- and low-cost munitions in
Desert Storm.)


--------------------
\24 All munitions costs are presented in 1991 dollars. 

\25 Depending on the platforms involved, the delivery of unguided
munitions would (in some cases but not all) require more aircraft
sorties than would the delivery of guided munitions.  This would
increase the cost of the unguided delivery, and it would expose a
larger number of aircraft to defenses.  However, guided munition
delivery requires more straight and predictable flight time and
greater pilot workload, thus making guided munition aircraft
vulnerable to defenses.  In short, the cost and survivability
trade-offs between guided and unguided munitions are not simple, and
the cost difference, if any, can be assessed only on the basis of
specific delivery circumstances. 

\26 These implications need to be considered within a wider array of
issues not discussed here, such as delivery platform cost and
survivability as well as munition capabilities and effectiveness. 


      ACHIEVEMENT OF CAMPAIGN
      OBJECTIVES BY AIR POWER
---------------------------------------------------------- Letter :5.3

Air power was clearly instrumental to the success of Desert Storm,
yet air power achieved only some of its objectives, and clearly fell
short of fully achieving others.  Even under generally favorable
conditions, the effects of air power were limited.  Some air war
planners hoped that the air war alone would cause the Iraqis to leave
Kuwait (not least by actively targeting the regime's political and
military elite), but after 38 days of nearly continuous bombardment,
a ground campaign was still deemed necessary. 

There were some dramatic successes in the air campaign.  It caused
the collapse of the national electric grid and damaged up to 80
percent of Iraq's oil-refining capacity.  At the end of the campaign,
only about 40 percent of the Iraqi air force survived. 

While air supremacy was achieved within the first week of the
campaign, delivery at low altitudes remained perilous throughout the
war because of the ever-present AAA and IR SAMs.  Iraq's C\3 and LOC
capabilities were partially degraded; although more than half of
these targets were successfully destroyed, Saddam Hussein was able to
direct and supply many Iraqi forces through the end of the air
campaign and even immediately after the war. 

Lack of intelligence about most Iraqi nuclear-related facilities
meant that only less than 15 percent were targeted.  The concerted
campaign to destroy mobile Scud launchers did not achieve any
confirmed kills.  Central Intelligence Agency (CIA) analysis showed
that more than 70 percent of the tanks in three Republican Guard
divisions located in the Kuwait theater of operations (KTO) remained
intact at the start of the ground campaign and that large numbers
were able to escape across the Euphrates River before the cease-fire. 
(Our assessment of the degree to which the objectives were achieved
is in app.  III; the development of the Desert Storm objectives is
described in app.  V.)


      FACTORS AFFECTING THE
      EFFECTIVENESS OF AIR POWER
---------------------------------------------------------- Letter :5.4


         SUCCESS ATTRIBUTABLE TO
         WEIGHT AND TYPE OF EFFORT
         EXPENDED
-------------------------------------------------------- Letter :5.4.1

The mix of available aircraft types enabled the United States and the
coalition to successfully attack or put pressure on a variety of
targets and target types; at various times of the day and night; in
urban, marine, and desert environments; with various guided and
unguided munitions.  Even including the platform and munition
preferences discussed above, no target category was exclusively
struck by a single type of aircraft, and no type of aircraft or
munition was exclusively used against a single type of target or
target category. 

Older, less costly, and less technologically advanced aircraft and
weapon systems made substantial contributions to the air campaign as
did the newer, more technologically advanced systems.\27 No
particular weapon system--whether of low or high technology, new or
old, single or multirole, high or low cost (or in between on any of
these criteria)--clearly proved more effective than another or
demonstrated a disproportionate contribution to the objectives of the
campaign.  For example, while the F-117 carried more tonnage per day
than the F-111F, the latter reported a higher rate of success hitting
the same targets using the same munitions; the F-16 had only a
slightly higher success rate than the F/A-18 when using the unguided
MK-84 against similar types of targets.  The B-52 and F-16 dropped
the largest known bomb tonnages, the F-16 and A-10 had the highest
sortie rates, and the B-52 and A-10 were cited by Iraqi prisoners of
war as the most feared of the coalition aircraft.  (The weight of
effort (WOE) and type of effort (TOE) that proved successful in the
air campaign are in apps.  II and VIII; specific weapon system
comparisons are in apps.  III and IV.)


--------------------
\27 The Desert Storm air campaign may have been the last large-scale
employment for several of the older types of aircraft.  For example,
the A-6E fleet is scheduled to be retired by 1998; the F-4G and F-111
fleets by fiscal year 1997; and all but two wings of the A-10 fleet
by the end of fiscal year 1996. 


         INTELLIGENCE NEEDS NOT
         FULLY MET
-------------------------------------------------------- Letter :5.4.2

Intelligence shortfalls led to an inefficient use of guided and
unguided munitions in some cases and a reduced level of success
against some target categories.  The lack of sufficient or timely
intelligence to conduct BDA led to the additional costs and risks
stemming from possibly unnecessary restrikes.  For example, BDA was
performed on only 41 percent of the strategic targets in our
analysis.  Restrikes were ordered to increase the probability that
target objectives would be achieved.  This may partly account for the
high tonnage of munitions expended on strategic targets--averaging
more than 11 and 44 tons of guided and unguided munitions,
respectively, for successful outcomes and more than 9 and 53 tons of
guided and unguided munitions, respectively, for less than fully
successful outcomes. 

Insufficient intelligence on the existence and location of targets
also inhibited the coalition's ability to perform necessary strikes
and achieve campaign goals.  The lack of target intelligence meant
that [DELETED] major Iraqi nuclear-related installations were neither
identified nor targeted, and no mobile Scud launchers were
definitively known to have been located and destroyed.  (See apps.  I
and III.)


         LIMITATIONS IN TARGET
         SENSORS INHIBITED
         EFFECTIVENESS
-------------------------------------------------------- Letter :5.4.3

The capabilities of target location and acquisition sensors were
critical to the effectiveness and efficiency of the air campaign.  IR
sensors allowed night operations, and although pilots praised many
sensor systems, they also pointed out numerous shortcomings.  IR, EO,
and laser systems were all seriously degraded by weather conditions
such as clouds, rain, fog, and even haze and humidity.  They were
also impeded by dust and smoke.  At high altitudes and even at low
altitudes in the presence of high humidity or other impediments,
pilots were unable to discriminate targets effectively.  They
reported being unable to discern whether a presumed target was a tank
or a truck and whether it had already been hit by a previous attack. 

Radar systems were less affected by weather, but the poor resolution
of some radars made it impossible to identify targets except by
recognizing nearby large-scale landmarks or by navigating to where
the target was presumed to be.  Radar systems specifically designed
for target discrimination and identification suffered reduced
resolution at the higher altitudes (and greater standoff distances)
where they were operating.  Pilots told us that the F-15E's
high-resolution radar, while designed to detect an object as small as
[DELETED] at a distance of [DELETED], could actually discriminate
only between a tank and a car at a range of about [DELETED].  (Target
identification and weapon system sensor issues are discussed in app. 
II.)


         CAMPAIGN PLANNING FAILED
         TO ANTICIPATE THE BDA
         LIMITATIONS
-------------------------------------------------------- Letter :5.4.4

The kinds of constraints encountered in Desert Storm do not appear to
have been adequately anticipated in planning the air campaign.  The
air campaign planners were overoptimistic concerning the number of
days that each phase of the campaign would require and the level of
damage each objective would require.  Moreover, many of the early
missions were canceled because of adverse weather, and after the
initial strikes were conducted, the BDA was neither as timely nor as
complete as planners had apparently assumed it would be. 


         CONTRIBUTIONS AND
         LIMITATIONS OF ADVANCED
         TECHNOLOGIES
-------------------------------------------------------- Letter :5.4.5

Desert Storm demonstrated that many newer systems incorporating
advanced technologies require specific operating conditions for their
effective use.  However, these conditions were not consistently
encountered in Desert Storm and cannot be assumed in future
contingencies.  Therefore, the level of success attained by various
costly and technologically advanced systems in Desert Storm may not
be replicated where conditions inhibit operations even more. 

Although much of what has been written about Desert Storm has
emphasized advanced technologies, many of these were subject to
significant operating constraints and a lack of flexibility that
limited their contributions and effectiveness.  [DELETED] While the
TLAM risks no pilot, it achieved a hit rate that CNA and DIA
estimated at [DELETED] percent, and it is costly.  [DELETED]
(Limitations on weapon system performance are discussed in app.  II.)

These limitations need to be recognized and anticipated when planning
air strikes or estimating the likely effectiveness of air
power--particularly for a short conflict, when there may not be
opportunities to restrike missed or partially damaged targets.  Even
in Desert Storm--with months of planning and a vast array of
in-theater resources available from the very start--uncertainties and
unknowns were typical rather than the exception. 


      DESERT STORM'S UNIQUENESS
      LIMITS LESSONS LEARNED
---------------------------------------------------------- Letter :5.5

The relevance of the air campaign in Desert Storm to likely future
contingencies depends at least partially on how closely its operating
conditions can be judged to be representative of future conditions. 
In this respect, Desert Storm's lessons are limited in some regard
because the environmental and military operating conditions for
aircraft and weapon system performance are unlikely to be repeated
outside southwest Asia and because future potential adversaries--not
least, Iraq itself--are likely to have learned a good deal about how
to reduce the effectiveness of guided weapons, such as LGBs.\28 At
the same time, performance in Desert Storm can be highly instructive
about the performance and outcomes that can be expected with existing
technologies under conditions like those encountered over Iraq. 


--------------------
\28 It is appropriate to note that "aggression by a remilitarized
Iraq against Kuwait and Saudi Arabia" was one of two scenarios
envisioned in planning strategy, force structure, and modernization
programs in DOD's BUR report. 


         COMBAT CONDITIONS OVER
         IRAQ AND KUWAIT
-------------------------------------------------------- Letter :5.5.1

The terrain and climate in Iraq and Kuwait were generally conducive
to the employment of air power.  The terrain was relatively flat and
featureless as well as devoid of vegetation that would obscure
targets.  Although the weather was the worst in that region in 14
years, weather conditions even less conducive to an air campaign
would be expected in many other locations of historic or topical
interest such as Eastern Europe, the Balkans, or North Korea.\29

(See app.  II.)


--------------------
\29 For example, the average percentage of time that the cloud
ceiling over Baghdad is less than or equal to 3,000 feet is,
historically, only 9 percent; comparable percentages over Beirut,
Lebanon; Osan Air Base, Korea; and St.  Petersburg, Russia; are 17,
33, and 64, respectively. 


         SIX-MONTH PERIOD TO
         DEPLOY, TRAIN, AND
         PREPARE FORCES
-------------------------------------------------------- Letter :5.5.2

The success of the air campaign is also attributable, in part, to the
6 months of planning, deployment, training, and
intelligence-gathering preceding Desert Storm.  During this interval,
President Bush assembled a coalition of nations that augmented U.S. 
resources and isolated Iraq.  War preparations were also aided by
preexisting facilities in the region and the lack of Iraqi
interventions to slow or deter the buildup of forces.  (See
app.  II.)


         SOME ENEMY CAPABILITIES
         OVERSTATED OR POORLY
         EMPLOYED
-------------------------------------------------------- Letter :5.5.3

Contrary to widespread prewar and postwar claims, the Iraqi IADS was
not "robust" or "state of the art." Rather, its computers were
limited in their capacity to monitor incoming threats; the system was
vulnerable to disruption by attacks on a relatively few key nodes;
and its design was [DELETED].  IADS had been designed to counter
limited threats from the east (Iran) and west (Israel), not an attack
from a coalition that included nearly 1,600 U.S.  combat aircraft
primarily from the south, hundreds of cruise missiles, and the most
advanced technologies in the world. 

On various dimensions, the Iraqi armed forces were not well disposed
to effectively counter the coalition's armed response to the Iraqi
seizure of Kuwait.  After U.S.  and coalition aircraft dominated
early air-to-air encounters, the Iraqi air force essentially chose to
avoid combat by fleeing to Iran and hiding its aircraft or putting
them in the midst of civilian areas off-limits to attack by coalition
aircraft.  Except for the failed Iraqi action directed at the town of
Khafji, the Iraqis did not take any ground offensive initiative
throughout the air campaign, and the coalition was able to repeatedly
attack targets, including those missed or insufficiently damaged on a
first strike.  As a result, when the ground war began, Iraqi ground
forces had been subjected to 38 days of nearly continuous
bombardment.  Evidence from intelligence analyses and prisoner-of-war
interviews also indicated that many Iraqi frontline troops had low
morale and were prone to heavy desertions even before the air
bombardment started. 

During the war, the Iraqis were unable to effectively resist
coalition air attacks from medium and high altitudes.  While the
Iraqis maintained a potent AAA and IR SAM threat to aircraft below
10,000 feet, the lack of an active Iraqi fighter threat (especially
after the first week); the coalition's suppression of most
radar-guided SAM defenses in the early days of the war; and the Iraqi
use of many of the remaining radar SAMs in an ineffective, nonradar
mode created a relative sanctuary for coalition aircraft at medium
and high altitudes.  Moreover, Iraq employed few potential
countermeasures (such as jamming) against coalition strikes.  (See
app.  II.)


         LIKELIHOOD OF VICTORY
         ALLOWED EMPHASIS ON
         SURVIVABILITY
-------------------------------------------------------- Letter :5.5.4

Given the overwhelming nature of the coalition's quantitative and
qualitative superiority, the conflict was highly asymmetric.  U.S. 
and coalition commanders controlled strike assets that were
numerically and technologically superior to the capabilities of the
enemy.  They expressed little doubt of a victory.  One result of this
was a command emphasis on aircraft and pilot survivability.  The
philosophy was "No Iraqi target was worth an allied pilot or
aircraft."\30

Other operating decisions were also taken to increase survivability. 
For example, after two F-16 losses on day three in the Baghdad area,
the Air Force ceased tasking large package daylight strikes of F-16s
against metropolitan Baghdad targets.  Similarly, after A-10 attacks
on the Republican Guard, during which two aircraft were hit while
operating at lower altitudes, the A-10s were ordered to cease such
attacks.  Instead, much higher altitude attacks by F-16s and B-52s,
with unguided bombs, were used.  (See apps.  II and III.)


--------------------
\30 GWAPS, Highlights (briefing slides), p.  30. 


      SOME AIRCRAFT AND WEAPON
      SYSTEM PERFORMANCE
      DIMENSIONS NOT TESTED
---------------------------------------------------------- Letter :5.6

A number of lessons cannot be drawn directly from Desert Storm
because systems were not stressed in ways that could be considered
likely and operationally realistic for future conflict.  For example: 
(1) with little or no Iraqi electronic countermeasures against U.S. 
munitions, airborne intelligence assets, or target identification and
acquisition sensors, no data were obtained on how these systems would
perform in the presence of such countermeasures; (2) with almost no
Iraqi air-to-air opposition for most of the war, many U.S.  aircraft
were also not exposed to these threats; and (3) many U.S.  weapons
were not delivered within the low-altitude parameters for which they
were designed, both platforms and munitions (thus, we do not know how
they would perform if delivered lower). 

However, precisely because of the advantages enjoyed by the
coalition, the problems that were encountered should be especially
noted.  These include the substantial amounts of unguided and guided
munitions that were used to achieve successful target outcomes and
the severe effect that the weather had on target identification and
designation sensors--some of which had earlier been described to the
Congress as capable in "all weather," "adverse weather," or "poor
weather." (See apps.  II-IV.) These problems should be considered as
warning signs about the effectiveness of various systems and
technologies under more stressful circumstances in the future. 


   CONCLUSIONS
------------------------------------------------------------ Letter :6

Operation Desert Storm was a highly successful and decisive military
operation.  The air campaign, which incurred minimal casualties while
effecting the collapse of the Iraqis' ability to resist, helped
liberate Kuwait and elicit Iraqi compliance with U.N.  resolutions. 

Our analysis of the air campaign against strategic targets revealed
several air power issues that should be planned for in the next
campaign.  First, the effectiveness of air power in Desert Storm was
inhibited by the aircraft sensors' inherent limitations in
identifying and acquiring targets and by DOD's failure to gather
intelligence on the existence or location of certain critical targets
and its inability to collect and disseminate timely BDA.  Pilots
noted that IR, EO, and laser systems were all seriously degraded by
clouds, rain, fog, smoke, and even high humidity, and the pilots
reported being unable to discern whether a presumed target was a tank
or a truck and whether it had already been destroyed.  The failure of
intelligence to identify certain targets precluded any opportunity
for the coalition to fully accomplish some of its objectives.  And
the reduced accuracies from medium and high altitudes and absence of
timely BDA led to higher costs, reduced effectiveness, and increased
risks from making unnecessary restrikes. 

Second, U.S.  commanders were able to favor medium- to high-altitude
strike tactics that maximized aircraft and pilot survivability,
rather than weapon system effectiveness.  This was because of early
and complete air superiority, a limited enemy response, and terrain
and climate conditions generally conducive to air strikes. 
Low-altitude munitions deliveries had been emphasized in prewar
training, but they were abandoned early.  The subsequent deliveries
from medium and high altitudes resulted in the use of sensors and
weapon systems at distances from targets that were not optimal for
their identification, acquisition, or accuracy.  Medium- and
high-altitude tactics also increased the exposure of aircraft sensors
to man-made and natural impediments to visibility. 

Third, the success of the sustained air campaign resulted from the
availability of a mix of strike and support assets.  Its substantial
weight of effort was made possible, in significant part, by the
variety and number of air-to-ground aircraft types from high-payload
bombers, such as the B-52, to PGM-capable platforms, such as the
stealthy F-117, to high-sortie-rate attack aircraft, such as the
A-10.  A range of target types, threat conditions, and tactical and
strategic objectives was best confronted with a mix of weapon systems
and strike and support assets with a range of capabilities. 

Fourth, despite often sharp contrasts in the unit cost of aircraft
platforms, it is inappropriate, given aircraft use, performance, and
effectiveness demonstrated in Desert Storm, to characterize higher
cost aircraft as generally more capable than lower cost aircraft.  In
some cases, the higher cost systems had the greater operating
limitations; in some other cases, the lower cost aircraft had the
same general limitations but performed at least as well; and in still
other cases, the data did not permit a differentiation.  (See app. 
IV.)

Fifth, the air campaign data did not validate the purported
efficiency or effectiveness of guided munitions, without
qualification.  "One-target, one-bomb" efficiency was not achieved. 
On average, more than 11 tons of guided and 44 tons of unguided
munitions were delivered on targets assessed as successfully
destroyed; still more tonnage of both was delivered against targets
where objectives were not fully met.  Large tonnages of munitions
were used against targets not only because of inaccuracy from high
altitudes but also because BDA data were lacking.  Although the
relative contribution of guided munitions in achieving target success
is unknowable, they did account for the bulk of munitions costs. 
Only 8 percent of the delivered munitions tonnage was guided, but at
a price that represented 84 percent of the total munitions cost. 
During Desert Storm, the ratio of guided-to-unguided munitions
delivered did not vary, indicating that the relative preferences
among these types of munitions did not change over the course of the
campaign.  More generally, Desert Storm demonstrated that many
systems incorporating complex or advanced technologies require
specific operating conditions to operate effectively.  These
conditions, however, were not consistently encountered in Desert
Storm and cannot be assumed in future contingencies. 

Four issues arise from these findings.  First, DOD's future ability
to conduct an efficient, effective, and comprehensive air campaign
will depend partly on its ability to enhance sensor capabilities,
particularly at medium altitudes and in adverse weather, in order to
identify valid targets and collect, analyze, and disseminate timely
BDA.  Second, a key parameter in future weapon systems design,
operational testing and evaluation, training, and doctrine will be
pilot and aircraft survivability.  Third, the scheduled retirement of
strike and attack aircraft such as the A-6E, F-111F, and most A-10s
will make Desert Storm's variety and number of aircraft unavailable
by the year 2000.  Fourth, the cost of guided munitions, their
intelligence requirements, and the limitations on their effectiveness
demonstrated in Desert Storm need to be considered by DOD and the
services as they determine the optimal future mix of guided and
unguided munitions. 

DOD and associated agencies have undertaken initiatives since the war
to address many, but not all, of the limitations of the air campaign
that we identified in our analysis, although we have not analyzed
each of these initiatives in this report.  First, DOD officials told
us that to address the Desert Storm BDA analysis and dissemination
shortcomings, they have

  created an organization to work out issues, consolidate national
     reporting, and provide leadership;

  developed DOD-wide doctrine, tactics, techniques, and procedures;

  established more rigorous and realistic BDA training and realistic
     exercises; and

  developed and deployed better means to disseminate BDA. 

DOD officials acknowledge that additional problems remain with
improving BDA timeliness and accuracy, developing nonlethal BDA
functional damage indicators (particularly for new weapons that
produce nontraditional effects), and cultivating intelligence sources
to identify and validate strategic targets.  Moreover, because timely
and accurate BDA is crucial for the efficient employment of high-cost
guided munitions (that is, for avoiding unnecessary restrikes), it is
important that acquisition plans for guided munitions take fully into
account actual BDA collection and dissemination capabilities before
making a final determination of the quantity of such munitions to be
acquired. 

Second, DOD officials told us that the most sophisticated targeting
sensors used in Desert Storm (which were available only in limited
quantities) have now been deployed on many more fighter aircraft,
thereby giving them a capability to deliver guided munitions. 
However, the same limitations exhibited by these advanced sensor and
targeting systems in Desert Storm--limited fields of view,
insufficient resolution for target discrimination at medium
altitudes, vulnerabilities to adverse weather, limited traverse
movement--remain today. 

Third, DOD officials told us that survivability is now being
emphasized in pilot training, service and joint doctrine, and weapon
system development.  Pilot training was modified immediately after
the air campaign to meet challenges such as medium-altitude
deliveries in a high AAA and IR SAM threat environment.  Service and
joint doctrine now reflects lessons learned in Desert Storm's
asymmetrical conflict.  Several fighter aircraft employment manuals
specifically incorporate the tactics that emphasized survivability in
the campaign.  DOD and service procurement plans include new
munitions with GPS guidance systems, justified in part by their
abilities to minimize the medium-altitude shortcomings and adverse
weather limitations of Desert Storm while maximizing pilot and
aircraft survivability. 

Fourth, DOD officials told us that although Desert Storm's successful
aircraft mix will not be available for the next contingency, DOD and
the services have made plans to maintain an inventory of aircraft
that they believe will be more flexible and effective in the future. 
Flexibility will be anticipated partly from the modernization of
existing multirole fighters to enable them to deliver guided
munitions (the aircraft systems being retired are single-role
platforms), and their effectiveness is expected to increase as new
and more accurate guided munitions are put in the field.  However, we
believe that strike aircraft modernization and munition procurement
plans that include increasing numbers and varieties of guided
munitions and the numbers of platforms capable of delivering them
require additional justification.\31


--------------------
\31 In Desert Storm, 229 U.S.  aircraft were capable of delivering
laser-guided munitions; in 1996, the expanded installation of LANTIRN
on F-15Es and block 40 F-16s will increase this capability within the
Air Force to approximately 500 platforms.  The services have bought
or are investing over $58 billion to acquire 33 different types of
guided munitions totaling over 300,000 units.  (See Weapons
Acquisition:  Precision Guided Munitions in Inventory, Production,
and Development (GAO/NSIAD-95-95, June 23, 1995.) Air Force plans
reveal that nearly 62 percent of all interdiction target types in a
major regional conflict in Iraq could be tasked to either guided or
unguided munitions today (1995) but that will fall to approximately
40 percent in 2002.  Concurrently, the percentage of targets to be
tasked to only guided munitions will increase from 19 percent in 1995
to nearly 43 percent in 2002. 


   RECOMMENDATIONS
------------------------------------------------------------ Letter :7

Desert Storm established a paradigm for asymmetrical post-Cold War
conflicts.  The coalition possessed quantitative and qualitative
superiority in aircraft, munitions, intelligence, personnel, support,
and doctrine.  It dictated when the conflict should start, where
operations should be conducted, when the conflict should end, and how
terms of the peace should read.  This paradigm--conflict where the
relative technological advantages for the U.S.  forces are high and
the acceptable level of risk or attrition for the U.S.  forces is
low--underlies the service modernization plans for strike aircraft
and munitions.  Actions on the following recommendations will help
ensure that high-cost munitions can be employed more efficiently at
lower risk to pilots and aircraft and that the future mix of guided
and unguided munitions is appropriate and cost-effective given the
threats, exigencies, and objectives of potential contingencies. 

1.  In light of the shortcomings of the sensors in Desert Storm, we
recommend that the Secretary of Defense analyze and identify DOD's
need to enhance the capabilities of existing and planned sensors to
effectively locate, discriminate, and acquire targets in varying
weather conditions and at different altitudes.  Furthermore, the
Secretary should ensure that any new sensors or enhancements of
existing ones are tested under fully realistic operational conditions
that are at least as stressful as the conditions that impeded
capabilities in Desert Storm. 

2.  In light of the shortcomings in BDA exhibited during Desert Storm
and BDA's importance to strike planning, the BDA problems that DOD
officials acknowledge continue today despite DOD postwar initiatives
need to be addressed.  These problems include timeliness, accuracy,
capacity, assessment of functional damage, and cultivation of
intelligence sources to identify and validate strategic targets.  We
recommend that the Secretary of Defense expand DOD's current efforts
to include such activities so that BDA problems can be fully
resolved. 

3.  In light of the quantities and mix of guided and unguided
munitions that proved successful in Desert Storm, the services'
increasing reliance on guided munitions to conduct asymmetrical
warfare may not be appropriate.  The Secretary should reconsider
DOD's proposed mix of guided and unguided munitions.  A reevaluation
is warranted based on Desert Storm experiences that demonstrated
limitations to the effectiveness of guided munitions, survivability
concerns of aircraft delivering these munitions, and circumstances
where less complex, less constrained unguided munitions proved
equally or more effective. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

The Department of Defense partially concurred with each of our three
recommendations.  In its response to a draft of this report, DOD did
not dispute our conclusions; rather, it reported that several
initiatives were underway that will rectify the shortcomings and
limitations demonstrated in Desert Storm.  Specifically, it cited (1)
the acquisition of improved and new PGMs, (2) two studies in
process--a Deep Attack/Weapons Mix Study (DAWMS) and a Precision
Strike Architecture study, and (3) several proposed fiscal year 1997
Advanced Concept Technology Demonstrations (ACTD) as programs capable
of correcting Desert Storm shortcomings.  In addition, DOD emphasized
the importance of providing funds to retain the operational test and
evaluation function to ensure the rigorous testing of our weapons and
weapon systems.  (See app.  XII for the full text of DOD's comments.)

We agree that the actions DOD cited address the shortcomings in
sensors, guided munitions, and battle damage assessment we report in
our conclusions.  However, the degree to which these initiatives are
effective can be determined only after rigorous operational test and
evaluation of both new and existing munitions and after the
recommendations resulting from the Deep Attack/Weapons Mix and
Precision Strike Architecture studies have been implemented and
evaluated.  Moreover, we concur with the continuing need for
operational test and evaluation and underscore the role of this
function in rectifying the shortcomings cited in this report. 

DOD also supplied us with a list of recommended technical
corrections.  Where appropriate, we have addressed these comments in
our report. 

If you have any questions or would like additional information,
please do not hesitate to call me at (202) 512-6153 or Kwai-Cheung
Chan, Director of Program Evaluation in Physical Systems Areas, at
(202) 512-3092.  Other major contributors to this report are listed
in appendix XIII. 

Joseph F.  Delfico
Acting Assistant Comptroller General