Download the dynamic market for remotely sensed data

THE DYNAMIC MARKET FOR REMOTELY SENSED DATA
Kass Green, President, Alta Vista
INTRODUCTION
Satellites and aircraft offer humans a view of their surroundings that humans cannot obtain on their own.
Well before the first man went aloft attached to a hot air-balloon in 1783, humans had long been fantasizing
about flight. From the Greek fable of Icarus, who flew too close to the sun, to Da Vinci’s Codex on the Flight of
Birds (1506) to our fascination with space flight today, humans have always envied birds and sought flight. Once
humans invented successful flying machines, it was an easy step to put cameras in flying machines so that others
besides the pilot could gain from the pilot’s perspective. While it took over 2000 years to develop both cameras
and airborne platforms, it took only 25 years for humans to take cameras into aircraft.
The innovation of air and space remote sensing has fundamentally changed the way we inventory and
manage resources, perform research, and respond to disasters. However, it is our conduct of war that has both
been most affected by remote sensing, and which has most effected remote sensing innovation. Before man
went aloft military intelligence could only be gathered from spies on the ground or by cavalry forays into enemy
territory – risky adventures that were not comprehensive and were prone to error and failure. Because there was
little ability to “see” behind enemy lines, armies could operate behind their own lines with almost complete
immunity.
Remote sensing provided both the ability to view the entire battlefield landscape at once, but more
importantly, to see behind battlefield lines. Movements of forces, construction of factories or the placement of
missile silos all became transparent to enemy personnel.
Wars also impacted innovation in remote sensing. For example, the need for camouflage arose to hide
troops and equipment from enemy remote sensing reconnaissance; and the development of near infra-red
multispectral and now hyperspectral imagery primarily arose from the need for remote sensing to allow
interpreters to distinguish vegetation from camouflage.
Gaspard Tournachon took the first aerial photography from a balloon in 1855. Four short years later
Napoleon III commanded Tournachon to gather aerial reconnaissance information for the decisive Battle of
Solferino during the Franco-Austrian War. Innovation in and growth of remote sensing technologies has always
been intertwined with the needs of militaries. While civilian remote sensing applications will always be
important, they are dwarfed by military expenditures even today.
This paper explores the role that military requirements play in remote sensing and discusses the successes
and failures in commercialization of the technology. We start with the high and low resolution systems that were
developed in response to military requirements. Next, the moderate resolution systems are considered. The
paper ends with a summary which compares and contrasts the systems adoption and commercialization.
AIRBORNE REMOTE SENSING
Commercial inventors designed and constructed the first airplanes and cameras. Neither was developed
because of a stated requirement from the military. However, when World War I (WWI) started, both sides in the
conflict quickly recognized the usefulness of reconnaissance photography from aircraft. As a result, military
requirements greatly advanced the combined uses of aircraft and camera technologies of remote sensing.
Cameras were placed inside planes rather than outside. Cameras and film were designed specifically for aerial
photography, and photo interpretation became a profession. The advancement somewhat slowed during the
period between the World Wars, but soon proceeded swiftly during World War II (WWII), again resulting in
better cameras, aircraft, and film. Following WWII, resources (personnel, cameras, and aircraft) quickly
migrated from the military to the commercial sector, as remote sensing was used to help (1) rebuild Europe and
Japan, and (2) support mapping of the rapidly growing American suburbs1. Surplus military aircraft and cameras
were abundant and affordable. Experienced pilots, camera operators, and photo interpreters were plentiful.
1
The technology was still referred to as aerial reconnaissance, as the term remote sensing did not appear until
the early 1960’s.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Advancements in remote sensing technologies were furthered by both the commercial and the public sectors,
often in partnership. The fixed costs of starting an aerial photography or mapping firm were relatively low.
As a result, civilian demand for remote sensing data exploded in both commercial and government markets.
Because a vital commercial industry existed which could support both private and civilian government needs, the
production of high resolution airborne imagery primarily became the purview of commercial companies who
provide imagery to both private and civilian public agency clients. Government airborne imagery capture is
reserved predominantly for research or for military or intelligence reconnaissance over areas where commercial
aircraft are prohibited2.
Today the US has over 180 mapping firms providing services and products to government agencies and
private entities. Even sensors designed and built with government research funds are often flown and operated
by private companies. The commercial airborne remote sensing industry is highly successful and highly
competitive.
SATELLITE REMOTE SENSING
At the conclusion of World War II, US troops captured both German rockets and rocket scientists. Both
were brought to White Sands, Arizona, and the first image of earth from space was taken from a captured V-2
rocket in 1949. As the US space missions evolved, a natural partition occurred between low spatial resolution3
satellites used for weather forecasting and high resolution systems used for reconnaissance. While the demand
for both originated with the military, access to the high resolution data was confined to the military for almost 40
years, while the weather data were immediately made accessible to the public. Conversely, moderate resolution
systems were developed for resource management. The requirements for the moderate resolution systems arose
from civilian researchers. Major military use of moderate resolution systems has only evolved in the last decade
and a half. This section reviews the evolution of all three types of satellite systems, the military’s impact on the
systems, and the various attempts to commercialize them.
Low Resolution Systems
Knowledge of weather is important for human survival, and the gathering of data about weather has been
supported by national governments for hundreds of years. After all, “The weather affects everyone every day,
and has an overwhelming influence on agricultural productivity and water resources, and thus on the economy of
all human communities” (Weiss and Buckland, 1997).
Like reconnaissance, the management of weather data collection started in the military because of the strong
impact weather has upon the conduct of war. US military weather reporting began in the early 19th century. In
1871 the US Congress created a weather bureau within the Army Signal Corps. It was transferred to the
Department of Agriculture in 1891, and in 1970 became part of the newly created National Oceanographic and
Atmospheric Agency (NOAA), which now operates all US civilian weather satellites.
The first earth observing system was a low resolution weather satellite named TIROS (Television and
Infrared Observation Satellite). The system was developed by RCA under a contract with the Army Ballistic
Missile Agency, and launched by NASA in 1960. Commercial and civilian government sectors quickly grasped
the value of satellite weather observations. From 1960 to 2004, over 40 civilian weather satellites were built and
launched by NASA and NOAA. Numerous weather satellites have also been launched by other countries.
Soon after the launch of TIROS, the military determined that it needed a separate weather satellite system to
support the strategic needs of the CORONA high resolution reconnaissance program (launched 4 months after
TIROS under the name Discover). They established the Defense Meteorological Satellite Program (DMSP)
which launched its first of multiple satellites in 1962. However, in 1994, the management of US weather
2
For example, The US government routinely uses Predator Unmanned Aerial Vehicles (UAVs) for
reconnaissance in the Middle East.
3
Spatial resolution is the ground size of the smallest picture element or “pixel” captured by the remote sensing
system. It is a function of the capability of the sensor and the altitude of the platform.. High spatial resolution
systems capture more detail than low resolution systems. There are no hard and fast definitions of low, moderate,
and high spatial resolutions. Typically, low resolution systems have a spatial resolution of 100 meters or above.
Moderate resolution systems are 5 to 99 meters, and high resolution systems have pixel resolutions of below 5
meters.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
satellites came full circle with the creation of the National Polar Orbiting Operational Environmental Satellite
System (NPOESS) which is a dual purpose program servicing both civilian and military needs by merging
NOAA’s Polar Orbiting Operational Environmental Satellite (POES) with the military’s DMSP (White House,
1994) .
The high cost of building and launching weather satellites coupled with the overwhelming public (military
and civilian) need for weather data, resulted in the building, launching, and management of low resolution
weather satellites remaining in the government’s hands. Weather data has been deemed a public good, which the
government decrees should be provided to the public at minimal expense. However, the use of the imagery and
information derived from weather satellites has quickly spread to other sectors as the imagery became ubiquitous,
appearing on our TV’s and now our computers continuously. By the late 1990’s, government procured and
disseminated satellite weather information provided the basis for a $200-$250 million a year commercial
industry (Weiss and Backlund, 1997). While the satellites have not been commercialized, the data from them has.
High Resolution Systems
The requirement for high resolution reconnaissance satellite imagery developed because of the need to
capture detailed images of areas where aircraft could not safely fly. Gary Power’s U2 reconnaissance jet was
shot down over the USSR in 1960. At that time, most countries still held to earlier treaties that stipulated that the
space over a country’s territories was as sovereign as the territories themselves. Peace time overhead
reconnaissance was considered to be a violation of a country’s sovereign rights, carrying some the same
implications as a ground invasion. The Russians and much of the world were incensed when Powers was
discovered spying. While the United States advocated the acceptance of an open skies policy in the United
Nations as early as 1958, the Soviets rejected it until 1986, after both the US and the USSR had launched
numerous spy satellites. Open skies policy is based on several UN legal principles which sanction satellite
remote sensing throughout the world without a sensed nation’s prior approval (United Nations, 1986).
From the 1960’s through the 1980’s spy satellites were pivotal sources of information for reconnaissance and
treaty verification during the Cold War. President Reagan’s tenet of “trust but verify” was embodied in the high
resolution space remote sensing programs. The US would trust its treaty partners, but also use reconnaissance
satellites to verify treaty adherence. The Soviets did the same.
Having better reconnaissance technology than your enemy or having a reconnaissance capability when your
enemy had none has been pivotal for national security. For example, during Desert Storm in 1991 the coalition
forces executed a “left hook” where 250,000 soldiers traveled to the west and north behind Iraqi lines which cut
off all Iraqi paths of retreat. By that time the coalition forces controlled the airspace and because the Iraqis had
no access to current satellite imagery, they were unable to “see” or prepare for the left hook. It was executed as
a complete surprise to the Iraqi forces. Understandably, the technology to create, launch and operate high
resolution satellite systems is considered an important national asset, protected by the highest levels of security
classification.
By the end of the Cold War, the large US aerospace corporations became concerned that the need for their
technology would become obsolete for reconnaissance purposes and that government expenditure for
reconnaissance missions would be greatly reduced (Jones, 2004). Additionally, American coalition partners in
theaters such as the first Gulf War were pressing for the US to share high resolution remotely sensed imagery
(Hays, 2001). Moderate resolution SPOT and Landsat were used throughout the war to provide large synoptic
views of operations. However, US partners were not allowed to view US high resolution reconnaissance imagery
because it was classified by the US government. Obtaining a non-classified source of high resolution imagery
became an imperative for sharing information among coalition partners. Finally, in 1992 the Russians
authorized the sale of film from one of their spy satellites which had a 2 meter resolution. Clearly, the US was
loosing the ability to control public access to worldwide high resolution satellite imagery.
In the early 1990s, Congress passed the 1992 Land Remote Sensing Policy Act (Public Law 102-555) which
authorized the development of commercial remote sensing satellite systems and directed the Department of
Commerce to regulate and license private sector parties to operate commercial systems. Buoyed by market
studies predicting a huge and fast growing market for remotely sensed data (Figure 1), US aerospace
corporations petitioned the administration to permit the commercialization of some of the technologies
developed for reconnaissance missions (Jones, 2004). The result was President Clinton’s Presidential Decision
Directive – 23, U.S. Policy on Foreign Access to U.S. Remote Sensing Capabilities (PDD-23) which allows the
export of remote sensing satellites, services, and high resolution imagery, but also retains the US government’s
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
right to suspend commercial operations (called shutter control 4 ) “during periods when national security or
international obligations and/or foreign policies may be compromised, as defined by The Secretary of Defense or
the Secretary of State…”. The directive is a compromise between a desire to have the US dominate the
commercial remote sensing market, the hope that the dominance would allow the US to control the proliferation
of remote sensing technologies, and the national security requirement of maintaining some sort of control over
the collection and distribution of high resolution imagery.
Early Projections of RS/GIS revenues
90
80
Hundreds of Millions of Dollars
70
60
NASA - 1995
50
SpaceVest - 1997
Frost and Sullivan 2000
40
Daratech
30
20
Figure 1. Early Projections
of Remote Sensing and
GIS revenues. All studies
report estimate total
projected revenues from
software, data, and services
in the Remote Sensing and
GIS markets. The Daratech
numbers also include
approximately $800
million for hardware.
10
19
89
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
0
Years
To date, over 30 licenses to launch and operate high resolution satellites have been granted by the
Department of Commerce. Three US companies (Space Imaging, OrbImage, and Digital Globe) have
successfully launched high resolution systems, with the highest resolution of one of them capable of sensing
objects down to .61 meters. Licenses for .5 meter systems have been granted and at least one company has
applied for a .25 meter license. In the mean time, other countries have continued to launch publicly available
higher and higher resolution systems, with the Israelis at 1.8, the French at 2.5 and the Indians at 5.8 meter
resolution.
Unfortunately, the civilian market for high resolution satellite imagery has not developed as projected, and
therefore the size of the market is much lower than indicated by earlier projections. Figure 2 is an update of
Figure 1 which shows remote sensing market projections performed in 2002 and 2003, almost a decade after the
4
So far, the US government has never invoked shutter control. The government recognizes that to do so could
seriously impact the ability of the commercial satellite imagery companies to establish a reliable market for their
products outside of the US government. However, immediately following 9/11, Space Imaging (at that time the
only commercial satellite imaging company with an operating satellite) voluntarily decided to sell imagery over
Afghanistan only to the US government. Subsequently the company entered into a contract with the National
Imagery and Mapping Agency (now the National Geospatial-intelligence Agency), to sell the company’s entire
satellite capacity over Afghanistan and portions of other middle eastern countries to the US government on a
month by month basis. This strategy effectively removed the commercial imagery of those areas from the
market without invoking shutter control. While often derided as “check book” shutter control by the press, the
solution was actually uniquely commercial. To control access to high resolution imagery over a specific area of
the earth at a specific time, the government merely bought all of the available satellite capacity, and thereby
avoided any first amendment arguments concerning “prior constraint” on the flow of information. Since this
contract, the government has declined to enter into similar contracts, looking to the companies to police
themselves, while still possessing the power to invoke shutter control should the need arise.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
heady days of the early 1990’s. Not only is the 2004 remote sensing market currently estimated to be less than
fifty percent of 1995 projections for 2004 (from $7 billion to $2.9 billion), but the projected rate of increase
(indicated by the slope of the lines) has fallen precipitously from estimates of 19-21% per year (KPMG, 1995)
to as low as 3.9% per year (Schutzberg, 2004).
Totall Projected Geospatial Revenues
90
80
Hundreds of Millions of Dollars
70
60
NASA - 1995
50
SpaceVest - 1997
Frost and Sullivan 2000
Daratech
40
Daratech
ASPRS - 2004
30
Figure 2. Total Projected
Geospatial Revenues.
Notice the large drop in the
Daratech estimates between
2001 and 2002
20
10
0
1989 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Year
One of the reasons the market for high resolution satellite imagery has not developed is because current
commercial satellite imagery products do not effectively compete with airborne imagery in areas like North
American and Europe, where a highly competitive airborne industry’s access to air is minimally restricted by
governments. While the costs of purchasing an airborne system are high ($500,000 – 3 million), they are
infinitesimal when compared to the costs building and launching of earth observing satellite systems ($200-700
million). Thus, airborne fixed costs5 are relatively low, allowing airborne imagery collection firms to offer nonlicense restricted imagery at prices that are comparable to those for licensed satellite imagery. In addition,
airborne imagery is also usually of higher spatial resolution and cloud free (satellite images are typically
guaranteed only to be no more than 20% clouded). Finally, satellites are tied to their orbits and designed scene
widths of their scenes, restricting when an image of a particular area can be captured. Conversely aircraft can fly
at any time, anywhere and underneath clouds. Transportation lines, coastlines, riparian corridors and
transmission lines will all be better captured by airborne systems. All told, the unrestricted airborne solution is
often of higher value than the licensed satellite solution. Understandably, the demand for commercial satellite
imagery tends to be focused in areas where commercial airborne systems do not operate (e.g. Iraq, North Korea,
etc) and from customers who are not particularly interested in sharing their data with a wide variety of users (e.g.
the National Geospatial-Intelligence Agency or NGA) 6 . If an airplane can gain access, airborne imagery
products will usually out-compete high resolution satellite products as they are currently configured.
Smaller than expected market size, lower rates of market growth, and the non-competitiveness of satellite
imagery products in much of the world have stagnated the ability of the commercial satellite companies to truly
commercialize – that is to be financially solvent without significant reliance on huge US military expenditures.
5
Fixed costs are the costs of inputs which are required to produce a product regardless of how many of the
products are produced. The costs of building and launching a satellite are fixed costs because they must be
borne whether 1 or 1 million images are captured following launch. Conversely, the variable costs required to
capture a satellite image are significantly less than the variable costs of capturing an airborne image. Once the
satellite is launched and operational, the cost to capture additional images is miniscule.
6
At least one commercial satellite company is beginning to move away from the restrictive licensing model for
imagery sales to US public agencies. The underlying assumption is that the increase volume of sales that would
result from dropping licensing restrictions would more than outweigh any loss in revenue from the sharing of the
data between agencies.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
The US military is the largest purchaser of commercial high resolution imagery, even though US government
expenditures on National Reconnaissance Office (NRO) systems dwarf their expenditures on commercial
satellite imagery. The commercial imagery compliments the NRO products, and helps to provide a safety net
should the well publicized problems with the NRO next generation of spy satellites persist (Pasternak, 2003).
Because imagery from the commercial satellites has become increasingly important to the military and
because the US government believes that our national security would be threatened by a loss of U.S. dominance
in this industry, President G. Bush released the U.S. Commercial Remote Sensing Policy in 2003 (White House,
2003). The policy’s goal is “to advance and protect U.S. national security and foreign policy interests by
maintaining the nation’s leadership in remote sensing space activities, and by sustaining and enhancing the U.S.
remote sensing industry.” Execution of the policy includes a directive that the U.S. government will “rely to the
maximum practical extent on U.S. commercial remote sensing space capabilities for filling imagery and
geospatial needs for military, intelligence, foreign policy, homeland security, and civil users…” NGA recently
assisted in the policy’s implementation with two programs, ClearView and NextView. ClearView provides for
NGA acquisition of commercial imagery to meet a portion of their mapping requirements. NGA has two
ClearView contracts, one with Space Imaging (approximately $120 million) and one with Digital Globe
(approximately $72 million) (Clark, 2004). The NextView program represents more of a partnership between the
government and the commercial companies by providing long term government commitments to purchase
imagery from the next generation of commercial high resolution systems. In exchange, NGA receives greater
control over access and priority of satellite tasking and the ability to influence the design of the new systems.
Two NextView contracts have been awarded, one to Digital Globe and one to OrbImage7. Both contracts have
an anticipated contract value of $500 million each over a span of 5 years. While these two contracts will
hopefully provide a strong enough base to keep both Digital Globe and OrbImage financially viable, they will
also result in increased influence of the military over the operations of commercial companies, thereby blurring
the line between military and commercial high resolution satellite operations.
In summary, the innovators of high resolution remote sensing satellites were the military, the early adopters
were the military, and the majority producers and users are still the military. Even though we now have a high
resolution commercial satellite industry, its primary purchasers are still, overwhelmingly, US and foreign
government militaries. As long as high resolution satellite imagery is less attractive than airborne imagery,
civilian buyers will continue to eschew the satellite data in favor of unlicensed, higher resolution and cloud free
airborne imagery. A future concern will be the possible impact of the highly subsidized commercial satellite
imagery on the competitiveness of unsubsidized commercial airborne systems.
Moderate Resolution Systems
The requirement for moderate resolution earth observing satellites was relatively slow to start, and,
significantly, was not generated by the military.8 It was not until the successful use of lunar observing satellites
by geologists to locate Moon landing sites in the 1960’s, that scientists within the Department of Interior and at
other government agencies conceived of a moderate resolution earth observing satellite for inventorying and
monitoring the earth’s resources. The resolution of the satellite needed to be high enough to allow for effective
monitoring and scientific study, but also low enough to not be perceived as a security threat by the US military.
Even with its coarse 80 meter resolution, the Earth Resources Technology Satellite (ERTS - later named Landsat
1) still initially alarmed many countries when it was launched in 1972.
The resulting 33-year Landsat program has been overwhelmingly successful. Landsat imagery is used for
multiple applications such as forest and agricultural monitoring, disaster planning and response, mineral
exploration, endangered species habitat identification, and famine relief among others. The Landsat Data
Continuity Mission (LCDM) (http://ldcm.usgs.gov/) web site summarizes the Landsat Program’s contribution
well; “Landsat represents the world's longest continuously acquired collection of space-based land remote
sensing data. For over 30 years, the Landsat series has collected and produced low-cost, moderate-resolution
multispectral data for researchers and decision-makers worldwide.”
7
The inability to secure a NextView contract by the third commercial high resolution satellite company, Space
Imaging, was devastating. That company is currently for sale. By the time of the Pecora 2005 meeting a sale
will probably be complete to one of three bidders. The likelihood of consolidation in the industry is high.
8
In fact, the military opposed the concept and argued instead that “the task of Earth resources survey can be
done easier, cheaper, sooner, and better, and in a politically more palatable and manageable manner from aircraft
than from satellites”(Katz, 1969).
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Adoption of Landsat imagery was slow at first. The resolution of the imagery from Landsats 1,2 and 3 was
too course for many applications, and while the multi-spectral character of the data facilitated automated image
classification, few people knew how or had computers powerful enough to conduct it. Adoption rapidly
accelerated with the improved resolution of subsequent launches (from 80 to 30 meters) and
ƒ
ƒ
ƒ
ƒ
Incorporation of GIS into image classification allowing for the integration of context and location into
image classification,
Development of exponentially faster computers with larger hard drives and memory,
Computer algorithms that increased the accuracy of terrain correction and imagery registration,
Development of robust image classification software.
Finally, the first Gulf War brought moderate resolution imagery (both the French SPOT and Landsat) into
the public’s homes every evening on the news as the media used the imagery to illustrate the progress of the war.
Over the last 15 years, the military use of Landsat has also increased. At the end of the Cold War, few could
visualize the unstable world we live in today, with terrorist attacks occurring at locations throughout the world.
The need for reconnaissance has not diminished as projected, but has greatly changed. Instead of focusing on
specific sites in one or two large countries (USSR and China), reconnaissance must now be dispersed around the
world, and new technologies are required. Since Desert Storm, the military has recognized the usefulness of
moderate resolution data to provide a wide synoptic view. In 1998 NASA contracted with Earthsat to create a
world wide mosaic of Landsat imagery (called GeoCover). NGA followed this program by partially funding the
development of Geocover LC which is a land cover and land use map that Earthsat created from the GeoCover
imagery. As a result, the military has become one of the biggest users of Landsat.
Even though the Landsat Program has been highly successful, it has been plagued by a lack of institutional
support. In 1979 President Carter shifted operation of the Landsat Program from NASA to NOAA and issued
Presidential Directive P-D 54 (White House, 1979) which stated that the goal of the federal government was
“eventual operation of these activities by the private sector”9. Five years later, President Reagan supported and
the Congress passed the Land Remote Sensing Commercialization Act of 1984 (PL 98-365) which directed
NOAA to migrate the Landsat Program from the federal government to the private sector. Subsequently a
competition was held and EOSAT, a joint venture of Hughes and GE, won the right to market, distribute, and
sell data from Landsats 4 and 5 with the government (NOAA) still bearing the operational costs of managing the
satellites. This included income from data sales, as well as from ground station operations worldwide. EOSAT
was to assume operational costs once the government-funded Landsat 6 was launched.
To fund the commercialization of Landsat, EOSAT increased the price of a Landsat data from approximately
$300 per scene to over $5000 per scene, which outraged most users – at that time mostly academicians and
civilian agencies. EOSAT also prohibited redistribution of the data, by selling a restricted license to use the data
rather than selling the data itself.
Over the next seven years the Landsat program remained a political football, tossed back and forth by
NOAA, EOSAT, the Administration, and Congress (Florini and Dehganzda, 1999). Finally after much review,
Congress passed the 1992 Land Remote Sensing Policy Act (Public Law 102-555) which ended Landsat
commercialization by designating that Landsat 7 would be taken back into the Federal government and managed
by a partnership of NASA and DoD. EOSAT retained the right to sell and distribute data from Landsats 4 and 5,
and Landsat 6 after it launched.
Within a year, the program again was in flux with the failure of Landsat 6 immediately after launch and by
DoD’s withdrawal of support for the program. In 1994 President Clinton signed Presidential Decision Directive
3 (White House, 1994) which gave NASA responsibility for developing and launching, and NOAA
responsibility for operating Landsat 7. Four years later, NOAA’s responsibilities were transferred to USGS.
Landsat 7 was successfully launched in April of 1999. The government set the price of a scene at $600
which they determined to be their cost of reproduction. The lower price of Landsat 7 imagery forced EOSAT
(by then owned by Space Imaging) to match the price on their sales of data from Landsats 4 and 5. Additionally,
the government dropped the ground station fees for downlinks from Landsat 7, which led many ground stations
9
In fact, President Carter’s suggestion was to commercialize all civilian earth observing satellite programs
(including the weather satellites)by transferring them from the government to the private sector. In March of
1983 the Reagan Administration moved forward with the transfer of both programs. However, Congressional
opposition stopped the transfer of the weather satellites.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
to abandon the much more expensive Landsat 5 ground stations. Unable to run Landsats 4 and 5 profitably,
Space Imaging returned their rights to distribute imagery from the satellites to the government in 2002.
In May of 2003, the Landsat 7 sensor (called ETM) experienced a partial but permanent failure of its scan
line corrector resulting in a loss of approximately 12% of each scene. While NASA and USGS have developed
methods for piecing together scenes from multiple dates to fill the gaps, the resulting product is insufficient for
some applications. Landsat 5 is still capturing excellent imagery even though it is over 20 years old; but because
its tape recorder failed over 10 years ago, Landsat 5 imagery cannot be captured over areas such as Russia or
Alaska that do not have Landsat ground receiving stations.
To compound the situation, in September of 2003 (three months after the Landsat 7 ETM+ failure), NASA
cancelled the request for proposals for the Landsat 7 follow-on mission (called the Landsat Data Continuity
Mission or LCDM) leaving the future continuity Landsat data in question. In August of 2004, the White House
issued a memorandum that directed USGS and NASA to initiate a partnership with the NPOESS Integrated
Program Office for the inclusion of a LDCM on NPOESS (Marburger, 2004), in effect, moving the management
of Landsat back to NOAA.
Over its 33 year history, the responsibility for the Landsat Program has been shifted from one organization to
another at least 5 times. That the program still exists in such a shaky institutional environment is astounding and
a testimony to the value of the data. Incorporating Landsat into a sustained operational environment such as
NPPOES will hopefully provide the Landsat Program with increased and sustainable institutional support.
NPPOES was designed to jointly support DOD and civilian weather requirements. Attaching Landsat to such a
well supported mission should increase the stability of the Landsat program. However, the launch of the
NPPOES satellite which is to carry LDCM is not scheduled until 2010, and many believe that the launch will be
delayed significantly beyond 2010 (partially because to change the NPPOES design to incorporate the Landsat
sensor is not a trivial task (ASPRS and MAPPS, 2005)).
Landsat 5 is now a very old system which could cease operations at any time. Landsat 7 imagery is seriously
compromised. Landsat data users, with time sensitive requirements, such as the Department of Agriculture,
currently face a gap in Landsat data continuity. No doubt the impact of the gap will spread to more users well
before NPPOES is launched.
In summary, the innovators of Landsat imagery were scientists and the early adopters were also scientists.
The majority users have been scientists and civilian agency personnel. Only recently has the user base expanded
and included the military and some commercial users. Advances in computers and software coupled with the
price reductions and the elimination of licensing restrictions following the launch of Landsat 7 have allowed
organizations to incorporate Landsat data into their day to day operations. NASA, DoD, USGS, USDA, NOAA,
BLM, FEMA, the National Park Service, the Forest Service, the Nature Conservancy, Conservation International,
the UN, Ducks Unlimited, and more than half of the fifty US states all rely on information derived from Landsat
data on a daily basis. In addition, several private companies, such as the Insurance Services Office, routinely use
Landsat data to produce commercial products. Programs dependent on Landsat imagery will not be able to
continue without the data because substitute imagery from other systems, such as the French SPOT, Indian IRS,
or NASA’s ASTER, is either too expensive or technically inadequate.
CONCLUSION
This paper concludes by addressing two questions:
1. Why did airborne remote sensing commercialize so quickly when both attempts to commercialize
satellite remote sensing have failed?
2. Why have the reconnaissance and weather satellite programs received consistent government support
while Landsat has struggled for most of its entire 33 year history?
Airborne verses Satellite Remote Sensing Commercialization
The pattern of remote sensing technology adoption and commercialization is highly influenced by whether
the system platform is space or air borne and the strength of the needs for the technology. Because of different
requirements and enormous disparity in fixed costs, airborne and satellite remote sensing technologies have
experienced very different patterns of adoption and commercialization. Following World War II resources
required for airborne remote sensing were abundant and relatively inexpensive. When matched to the huge
requirements for maps, the airborne market was bound to grow quickly. Even though the fixed costs of today’s
airborne digital camera systems can run well over $1 million, these costs are still dwarfed the several hundred
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
million dollar fixed costs of building, launching and operating an earth observing satellite. The fixed costs of
earth observing satellites creates a significant barrier to entry into the market that at this time can only be
overcome with government subsidies. The requirements of civilian users cannot support the investment
necessary to build, launch and operate an earth observing satellite. First, the size of the overall market is much
smaller than anticipated. But more importantly, satellite imagery cannot compete with airborne imagery in the
markets of advanced economies, which are where most of the non-military expenditure on remote sensing
originates.
The impact of the military on innovation and adoption of remote sensing technologies is clear. Military
requirements drive much of the innovation and adoption of remote sensing. Military strategic needs have also
constricted the release of remote sensing technologies into civilian markets. The technology to procure high
resolution imagery from space has been available to the US military for at least four decades, but available to
civilian users for only the last five years. The world we live in today, with abundant and minimally restricted
access to high resolution imagery worldwide, is relatively new and somewhat threatening to defense and
intelligence organizations. Given the roots of remote sensing technology in the requirements of the military, it is
perhaps surprising that the technology has been allowed to migrate to civilian markets at all. This nation’s
fundamental belief in democracy and capitalism and in the transparency upon which both are based, will always
create tension between the desire to conceal clandestine activities while simultaneously promoting open
information exchange.
Landsat Verses Weather and Reconnaissance Satellite Funding
Given that earth observing satellites require government subsidies, why have the reconnaissance and weather
satellite programs received consistent government support while Landsat has struggled for most of its entire 33
year history? First, both the reconnaissance and the weather satellites are needed by the military which provides
a deep well of support. The usefulness of Landsat data to the military has only recently been realized. Second,
the weather satellites also enjoy a broad base of support, which is why two parallel weather programs (military
and civilian) were funded for forty plus years. Each one of us relies on the weather satellites in some way on a
daily basis. Be it whether to bring an umbrella or not, or how to pack for a trip, or when to plant or harvest a
crop, or when to prepare food support for a nation in drought, information on the weather is important to
everyone which creates a broad base for weather information both inside and outside of the military. Finally,
unlike high resolution satellite imagery, the substitutes for weather satellites are significantly inferior. Before we
had weather satellites, we relied on reports from ships at sea and from planes. Before we had weather satellites,
we relied on reports from ships at sea and from planes and weather balloons. However planes and balloons
cannot fly high enough. Only satellites can provide the wide synoptic view of an entire weather system active
over a continent.
In contrast, Landsat’s base of support has been shallow, narrow, and fragmented. The user base was initially
made up of scientists and the difficulty of working with the data slowed its adoption for many years. Then four
years of ill conceived commercialization resulted in Landsat prices and licensing terms which stagnated the
adoption of the imagery. Without a strong user base, the institutional framework supporting Landsat has been
weak from the beginning. The program has bounced from one agency to another every five to eight years.
It has only been since 2000 that the use of Landsat data has broadened widely, and it is ironic that it is now
that the Program’s future is most at risk. The recent broadening of the user base may have come too late to
ensure the future continuity of Landsat data because the user base is not organized, nor is it politically active. As
with high resolution satellites, we may need to once again look to the military to save a vital remote sensing
program.
REFERENCES
The author wishes to thank Dr. Don Lauer for his thoughtful review of this paper and for insightful suggestions.
Alexander, T.M.. R.J. Birk, and D.P. Brannon. Analysis and Forecasts for the RS/GIS Industry and Markets:
1995-2005. Prepared for NASA Commercial Remote Sensing Program. NASA Stennis. 42pp.
ASPRS and MAPPS. 2005. American Society of Remote Sensing and Photogrammetry and Management
Association of Private Photogrammetric Surveyors. April 6, 2005. Letter to Dr. John Marburger,
Director, Office of Science and Technology Policy. Executive Office of the President.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Clark, Patrick. 2004. Commercial Satellite Imagery Matures as an Asset. in Military Geospatial Technology.
Volume 2 Issue 1.
Daratech. 2000. Press Release. GIS Software Growth 12.8% in ’99; Worldwide Software Revenues Top $845
Million. Daratech, Inc.
Daratech. 2001. Press Release. GIS Software Revenue Tops $1 Billion in 2001. Total User Spending for GISRelated Software, Hardware, Services Passes $7 Billion.. Daratech, Inc.
Daratech. 2004. Press Release. Worldwide GIS Revenue Forecast to Top$ 2.02 Billion in 2004, up 9.7% over
2003. Daratech, Inc.
Florini, A.M. and Y.A. Dehqanzada. 1999. “No more Secrets?” Policy Implications of Commercial Remote
Sensing Satellites” Washington D.C. Carnegie Endowment for International Peace. Carnegie Paper
Number 1, July 1999.
Frost and Sullivan. 2000. The GIS Data End User Marketplace: Trends, Issues, and Pricing. Presented at the
NASA Workshop for a NASA Earth Science Region Applications Initiative. Sacramento, CA. Sept.
11-13 2000.
Hays, Peter l. 2001. “Transparency, Stability, and Deception: Military Implications of Commercial High
Resolution Imaging Satellites in Theory and Practice”. Presented at the International Studies
Association Annual Congress, Chicago, Feb 21-24.
Jones, Dennis. 2004. Commercial Remote Sensing and National Security. In Crosslink. The Aerospace
Corporation magazine of advances in aerospace technology.
http://www.aero.org/publications/crosslink/summer2004/09.html
KPMG. 1994. MAPSAT Market Review. A KPMG report prepared for NASA.
Landsat Data Continuity Mission (LDCM) web site http://ldcm.usgs.gov/
Marburger, J. 2004. Landsat Data Continuity Policy. Executive Office of the President. Office of Science and
Technology Policy. August 13, 2004.
Mondello, C., G.F. Hepner, and R.A. Williamson. 2004. 10-Year Industry Forecast. Phase I-III – Study
Documentation. Photogrammetric Engineering & Remote Sensing. Volume 70. Number1. pg 5-58.
Pasternak, Douglas. 2003. Lack of Intelligence. U.S. News and World Report. August 11, 2003. 35-43.
Public Law 98-365. 1984. Land Remote Sensing Commercialization Act of 1984. July 17, 1984
Public Law 102-555. 1992. 1992 Land Remote Sensing Policy Act.
Schutzberg, A. 2004. The State of Commercial Remote Sensing Market: A Conversation with Ron Stearns of
Frost and Sullivan. Earth Observing Magazine. Volume 13. Issue 5. pg 33-34.
United Nations, 1986. "Principles Relating to Remote Sensing of the Earth from Outer Space," Principle XII,
UN General Assembly Resolution 41/65.
Weiss, P.N. and P. Blacklund. 1997. International Information Policy in Conflict: Open and Unrestricted Access
versus Government Commercialization,’ in Borders in Cyberspace, Kahin and Nesson editors. MIT
Press.
The White House. Office of the Press Secretary November 16, 1979. Presidential Directive 54 Civil
Operational Remote Sensing.
The White House, Office of the Press Secretary. May 10, 1994. Fact Sheet - Landsat Remote Sensing Strategy.
The White House, Office of the Press Secretary. March 10, 1994. Fact Sheet - Foreign Access To Remote
Sensing Space Capabilities
The White House, Office of the Press Secretary. May 10, 1994 Fact Sheet - Convergence of U.S. Polar Orbiting
Operational Environmental Satellite Systems
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota