Nancy Tosta
Director of Forecasting and Growth Strategy
Puget Sound Regional Council, Seattle, Washington
ntosta@seanet.com
Preface:
This article has originally appeared in the GeoInfo Systems Magazine and is
reprinted with permission of the author.
Geo Info Systems Vol. 7:5, May 1997: Data Data Column
A few weeks ago I had the opportunity to attend the GIS/GPS'97 Conference in Qatar and there
discovered - somewhat to my surprise - a National Spatial Data Infrastructure (NSDI). I should
have known. Zul Jiwani, the Canadian leader of the Qatar's Centre for Geographic Information
Systems (CGIS, Doha), had been telling me for years about data coordination in that country. I'd
also heard about the 1992 Exemplary Systems (URISA, Washington, D.C.) and 1995 Highest
Technological Achievement (ESRI, Redlands, California) awards that the CGIS had received.
But Zul - out of politeness, I suppose, knowing that we were going through such pains in the
United States - had never called it a "national spatial data infrastructure". After my years of
pushing NSDI around the United States, seeing the principles operating in Qatar was quite a
revelation. High resolution, fully integrated GIS data are available to all agencies and the public.
The agencies make use of the same base information, so all new data collected are also
integrated.
I've been thinking about the factors that contributed to Qatar's success and what they might
reveal about NSDI efforts in the United States, as well as those elsewhere in the world.
MIGHTY MICROCOSM
Although Qatar is a country, it is referred to as the "State of Qatar". It sits on the east coast of
Saudi Arabia, on the Persian Gulf. Few countries in the world are as small as Qatar, which has an
area of roughly 4247 square miles, somewhere between the size of the states of Delaware and
Connecticut (Wright, 1996). Most of the country is desert; when I asked how much it rained -
being from Seattle, one always wonders - I was told, "five or six hours per year."
Qatar's population in 1995 was about 641,000, equivalent to the 1995 population of North
Dakota (Qatar Central Statistical Organization, 1996). Every state in the United States except
three (Vermont, Wyoming, and Alaska) has more population than Qatar; so do 14 cities and 62
metropolitan statistical areas (Hoover's, Inc. 1996). Another statistic from Qatar's Central
Statistical Organization (1996) also provides a perspective on the country's size and rate of
development: there are 13 hotels countrywide.
Qatar has relatively few institutions. The national or "state" government includes the typical
number of agencies - maybe 20 - that you'd expect to find in an environment that size. There are
ten municipalities, but as far as I am aware, no government infrastructure associated with them.
There are essentially three cities: Doha, the capital, with a population of slightly more than
200,000 and a sleepy city government; Rayyan, with a population just over 91,000; and Wakrah,
with a population under 25,000 (Wright, 1996). Qatar is a traditional monarchy ruled by an emir,
who can issue mandates that are to be obeyed. The economy is dominated by oil, banking, and
shipping services (Wright, 1996). Their gross national product is 0.1 percent of the United
States's (Wright, 1996)
GIS CITY
One of the amazing sights at the Qatar GIS/GPS'97 Conference was the "GIS City". Most GIS
conferences, whether state, national, or international, sponsor an exhibit of GIS software,
hardware, and service providers. Qatar's exhibit was a simulation of a networked city depicting
the linkages among the Qatari agencies sharing data and GIS resources. All agencies using GIS in
Qatar are linked with GISnet, which is a dedicated 100 megabits-per-second FDDI fiber optic
network. The public can also access GIS data via this network.
The underlying data "infrastructure" that supports Qatar's GIS efforts was depicted at the
entrance to GIS City. The digital topographic database comprises 6,000 horizontal control survey
monuments for the country and 4,500 vertical control stations; digital orthoimagery (10-
centimeter pixels in urban areas, 1 meter elsewhere); digital elevation models (again, 10-centimeter vertical accuracy in urban areas, 1 meter elsewhere); and an impressive collection of
vector layers (for example, streets, buildings, zoning, land use, soils, utilities) at 1:1000 scale in
urban areas and 1:10,000 for the rest of the country. These databases are used by all of the
national agencies that have a need for geospatial data, including agriculture, environment,
planning, health, transportation departments - 16 in all.
In an interview with Zul Jiwani, I learned that CGIS is the hub for coordination among the
agencies. CGIS has an in-house staff of 67, more than 50 others who work for other agencies but
train or operate out of the center, and a budget of approximately US $1 million, not including
staff salaries. CGIS sets standards, provides training, supports the National GIS Steering
Committee, creates data sets, and in general coordinates all aspects of GIS development and use
throughout Qatar. Agencies, utilities, and municipalities are allowed to develop their own GIS,
but must comply with national standards.
The entire exhibit hall was set up to depict the interrelationships among these agencies and their
shared data base. Two of the more fascinating exhibits were the Central Statistical Organization
because it was about to undertake an incredibly comprehensive census on March 17, and the
Geographic Names Project, dedicated to the historic preservation of Qatar place names.
The availability of digital parcel and building data for the entire country allows the Central
Statistical Organization to conduct a General Housing and Population Census by literally
knocking on the door of every house. Census-takers have maps of all of the buildings in their
zones and are not expected to turn in their results until they can show that they visited every home
and counted the inhabitants, or have an explanation as to why that was not possible.
The Geographic Names Project has developed an audio and video history for Qatar through
interviews with elders who remember the origins of many of the names of landmarks and
landscape features across the country. The histories are geocoded to the locations of specifically
named landmarks. A sample of the types of features - with corresponding interviews - was
depicted in GIS City, in a small-scale model of the desert.
These exhibits were just two examples of Qatar's dedication to data collection. Other exhibitors
included the Departments of Environment, Health, Roads and Traffic, Urban and Regional
Planning, Drainage, Fisheries, Electricity and Water, Land Registration, Agriculture, and Police,
as well as Q-Tel, the Qatar Public Telecommunications Corporation (all of these agencies are
located in Doha)..
Qatar's GIS activities began in 1988 when Sheikh Ahmed bin Hamad Al-Thani, then head of the
Ministry of Municipal Affairs and Agriculture, witnessed a GIS demonstration in Canada. Al-Thani realized that this was a tool that could change how information was managed in Qatar and
became the project's champion, sharing his vision with cabinet level and senior members of
Qatar's government. After conducting a user needs assessment, the country was on its way.
There were no major existing GIS activities to compete with Qatar's efforts to coordinate and
implement GIS from the top-down. In October 1990, a National Steering Committee of high
level government representatives was formed and The Center for GIS was created. These
organizations have played critical roles since in conducting pilot studies, guiding standards
development, ensuring that standards are implemented in all agencies, and overseeing the
development of coordinated data sets.
NOT ALL NSDIs ARE ALIKE
I provided background statistics earlier to help establish the monumental differences in levels of
complexity - geographically, socially, and institutionally - between the State of Qatar and the
United States. Geography matters in spatial data infrastructures. The larger the area and the
more development, the greater the expense to build data sets. The more development, the greater
the likelihood that high resolution data will be needed for decision-making. The more institutions
and players, the greater the time commitment in coordination efforts. The more existing GIS
activity, the greater the difficulties in GIS coordination.
Fewer resources create more coordination challenges. This may seem contrary to common sense,
as it's often argued that less funding makes coordination more necessary. But in my experience,
most agencies strapped for funds will curtail activities with other agencies before eliminating
internal support programs. Therefore, adequate funding can enhance coordination, unless excess
funding - particularly if it's distributed inequitably - interferes, and the wealthy agency does what
it wants.
In my mind, the following factors made nationwide GIS coordination possible in Qatar:
It wasn't just that one of these factors existed; I would argue they were all necessary to one
degree or another. How does the United States - or for that matter, any other country - compare?
Let's go down the list.
Highest level political support. Yes, to some extent. President Clinton did sign Executive
Order #12906 mandating the creation of the NSDI through various federal actions. And
Secretary of the Interior Babbitt does chair the Federal Geographic Data Committee (FGDC).
But no other Secretaries sit on the FGDC, and many agency representatives are not "highest"
level.
Political will and authority to mandate and enforce standards. Not exactly. Although
Secretary Babbitt chairs the FGDC, not even his own agencies in the Department are totally
committed to carrying out the actions of the Executive Order. Many of the other FGDC
representatives do not have the clout in their agencies to affect change. It's not that they don't
care or make an effort, it's just that the federal government is an enormous organization with lots
of players and even more agendas.
Some agencies have made a valiant effort to require metadata, clearinghouse support, and
standards development. But other agencies that have collected data a certain way for decades do
not like to be told that they have to change, particularly if they are tracking patterns or trends
based on the data. The Office of Management and Budget (OMB) may occasionally make noise
about reducing funding if standards aren't followed, but it's not difficult for an agency to hide
programs. And because Congress actually passes the budget, OMB's attempts to enforce
standards may be lost on Capitol Hill. No federal agency is likely to acknowledge that what it
does duplicates or conflicts with another agency's activities, and no federal agency will be willing
to halt or eliminate programs because they replicate efforts at other agencies. Then again, this
isn't any more likely to occur at the state or local level.
Agencies have and do continue to reconcile differences and redundancies in data collection, but
this is usually a tedious, very long term process with meetings, meetings, and more meetings.
Sometimes it works in the end, and sometimes the players keep changing and the meetings keep
going. Besides, for the most part, the federal government cannot control state or local
government actions. No standards adopted by the federal government are mandatory at other
levels of government.
Outstanding technical leadership. Who do thousands of U.S. GIS users look to for guidance
when developing their GIS? Usually they'll start with their vendor; sometimes they'll look to the
federal government. There are many federal geospatial data base managers, and in some areas -
for example, image processing and large database management - probably more experts than
almost anywhere else in the world. Some of the classified work surely pushes the envelope in the
use of geospatial data. But, many of the agencies that are responsible for leading the development
of data standards under the umbrella of the FGDC are not major users of geospatial data. For
example, the United States Geological Survey (USGS) creates 7.5-minute quadrangles and digital
data bases derived from those quads, but aside from the USGS's Water Resources Division, the
agency is not a major user of its own or other geospatial data. For this reason, the FGDC has
worked to create committees and a review process for standards to encourage involvement from
state and local governments and private- and academic-sector representatives. These committees
provide some level of technical leadership.
Small and relatively uncomplicated geography. Hardly! If nothing else, the United States is
big and complicated: more than 850 times the size of Qatar, encompassing nearly every
ecosystem imaginable - with the possible exception of various tropical rain forests. More than
144,000 square miles are estimated to be developed in urban and transportation uses.
Few institutions. No, again. The population of the US as of July 1995 was 262,755,000. We
the people have formed a lot of institutions. The federal government in 1995 employed 2,857,371
people (non-military), while state and local governments employed roughly 16,457,000. There
are 13 major executive departments in the federal government, with as many as 90 separate
agencies and bureaus under them. Additionally, there are over 60 independent agencies such as
the National Aeronautics and Space Administration and the Environmental Protection Agency.
Each of the 50 states has tens or hundreds of agencies, as do most of the 3000 counties, 19,000
municipalities, and 16,000 townships, along with nearly 46,000 school and special districts. All
told, there are more than 85,000 units of government in the US, not counting all of the agencies
that make up those governments.
Adequate funding. Who knows? Given the level of institutional and geographic complexity in
the United States, what's adequate? The CGIS in Qatar expended approximately $5 million to
build the base data sets previously described. Additional funds were expended by each agency to
build other specific data sets. In comparison, a 1993 OMB study reported that the federal
government spends more than $4 billion annually on the collection and management of geospatial
data. This is spread across hundreds of programs, and with the possible exception of certain
military programs almost never exceeds $10 million per program.
Additionally, state and local governments have made significant investments in geospatial data.
My guess is that after federal agencies, local governments are the next biggest geospatial data-GIS spenders. A quick survey of six counties here in the Pacific Northwest revealed collective
expenses of at least $40 million to build their GISs and $5 million annually for maintenance.
Little existing GIS activity. No, the United States does not meet this criteria. There are
thousands of GIS efforts in the US. No one knows how many, but it's unusual to find an agency
or jurisdiction (unless it's so small as to not even have a computer) not using or contemplating the
use of GIS. Most jurisdictions starting their GIS efforts face the challenge of either relying on
existing data such as Census TIGER files or (more likely), building a data base that requires that
data standards be addressed. These jurisdictions will often look to other agencies to find out
what's been used before, or to a vendor or federal agency. They may copy what's been done
elsewhere or simply decide what works for them and proceed on that basis.
Once an organization has invested resources in a data base, its willingness to change formats or
adopt new standards drops rapidly. New standards are likely to entail a lot of work - with an
often less-than-optimal end result over what is currently being used. Convincing funding
authorities that a database needs to be rebuilt to meet outside standards may be difficult. And at
the local level, where most data must be generated anew rather than derived from others, there's
little incentive to standardize to meet other agency's needs.
THE REALM OF POSSIBILITY
So where does all this leave us here in the United States? Other nations may be smaller than the
United States, less complex geographically or institutionally, and have less GIS activity, hence
"scoring higher" with Qatar's criteria. But I can't think of one that would come close to meeting
the conditions in Qatar. Based on this totally arbitrary assessment of the conditions that
contributed to Qatar's success - but might be limiting in other situations - I've developed a rough
graph.
Figure 1 depicts a relationship among these criteria. The y axis depicts a variable that describes
the geography, both in terms of size and complexity. The x axis is a scale of the number of
institutions using GIS (and most likely generating data). The area under the curve (the shape of
which is purely hypothetical and has generated much argument among friends), defines the
environment where spatial data coordination is most likely to succeed. This region would be
where one might consider developing a spatial data infrastructure. The curve can be pushed out
by several factors, including resources (throw a lot of money into coordination), technological
superiority (provide a common tool or approach that everyone will want to use), political will
(have the ability to tell everyone what to do), or shared purpose (a common issue unites many
interests). These factors may make it possible to build an infrastructure over a larger area with
more players.
I can't tell you exactly what the numbers are on the axes, where geography is too big or too
complex, or how many existing GIS installations is too many. My feeling is that what's been done
in Qatar is not likely to be replicated in very many other environments around the world, at least
not at the national level. Qatar, to it's credit, identified the data needed for decision-making at all
levels, including where most decisions are made - locally - and developed the highest resolution
data to support these types of applications. Their small size allowed them to consider this a
national spatial data infrastructure.
Other nations - such as the United Kingdom, with an area less than the state of Oregon; Portugal,
about as large as Maine or Indiana; The Netherlands, slightly larger than Maryland; or even Japan,
just slightly larger than Montana - are working on or have indicated an interest in a 'national
spatial data infrastructure (Hoover's, Inc., 1996 and Wright, 1996). Based on their size, they
might accomplish some level of national coordination. However, with the exception of The
Netherlands, these nations have taken top-down approaches, one to such an extent that the
national mapping agency is seen as the only valid collector of data for the nation, charging high
fees for data access.
In the United States, several states and various metropolitan regions are on their way to
accomplishing something similar to Qatar's NSDI. These are not national spatial data
infrastructures, but they cover areas as large or larger than Qatar, involve more existing GIS
users, and are building data sets that are useful for decision-making.
The unit of geography that forms the basis for the infrastructure is as likely to be driven by
geography - such as a watershed or major urban area - as by a geographically arbitrary institution
such as a state. A common interest (such as transportation planning, water quality protection, or
ecosystem management) is more likely to smooth the way towards agreement on database
accuracy and other data characteristics than a mandate requiring coordination. These efforts are
not usually top down, but are the result of a cooperative effort among various levels of
government and different institutions that share common geographic interests and have
recognized the value of sharing and integrating data. These are often the agencies and institutions
that not only must share data, but must reach a consensus on practices and policies to manage
resources. Their ability to work together to share data may very well contribute to their ability to
work together on bigger issues.
The FGDC recognized long ago that the challenge in building a national spatial data infrastructure
in the United States was not necessarily uniformity, but rather consistency. That everyone will
use the same data sets in the United States, like they do in Qatar, is inconceivable. If nothing else,
we are too 'sophisticated' in our GIS development - too far gone may be another way to say it.
The FGDC has encouraged partnerships at the local level with incentive funding and is also
addressing development of basic guidelines for building data sets. These guidelines or standards
will encourage local high resolution data bases to contain common features to allow integration or
generalization to build larger-area data sets from multiple jurisdictions. The FGDC is still
working on these standards. Though it may not be coming from the highest levels of authority as
it did in Qatar, the FGDC has been a critical force in raising awareness about data-sharing
possibilities in the United States. Perfection is unlikely.
THINKING GLOBALLY, ACTING LOCALLY
Some argue that as more nations consider development of spatial data infrastructures we should
address the possibility of a global spatial data infrastructure. I certainly recognize and appreciate
that the world has become a global economy and many issues are global in nature. But I take
exception to expending many resources to build purely global data bases, at least beyond those
that have been generated or are being upgraded by agencies such as the Defense Mapping Agency
(now part of the National Imagery and Mapping Agency). Proponents of the GSDI argue that
many decisions, perhaps especially those that address environmental and sustainable development
issues can be addressed by global data sets. I don't buy it. Although, global sustainability may be
of concern - and data that allow modeling of global climate change and air pollution are critical -
it is because millions of local decisions are made - develop a wetland, let a field lie fallow, build
another subdivision, construct another freeway, pave another meadow - that sustainability is in
question. The cumulative effect of innumerable local decisions made without the value of good
information is changing the world. Is it more important to direct our limited resources for
geospatial data globally or locally? Perhaps the global question is more one of how we can best
network our local databases.
Meanwhile, something is growing in the United States that does not exist in most other parts of
the world: an incredibly rich network of data development experts at the local level who are
beginning to understand the value of geospatial data for sound decision-making. The United
States will probably never accomplish what Qatar did, but maybe there's another way to look at
it. We could have 850 times the number of people caring about and using geospatial data for
decision-making as Qatar does. We are beginning to put in place complex, robust, integrated
geospatial data systems that will ultimately affect every land use decision made in the United
States. Hats off to Qatar for their national spatial data infrastructure - I had an enlightening,
exciting, and wonderful visit. At the same time I'm humbled by the efforts of the thousands of
dedicated local database builders in the United States. There's not a lot of glory in what they do,
but there is value.
REFERENCES
Hoover's, Inc. 1996. The American Almanac 1996-97. Austin, Texas: Hoover's Inc.
Qatar Central Statistical Organization. 1996. Qatar in Figures. Doha, Qatar: Qatar Central
Statistical Organization.
Wright, J.W., ed. 1996. The Universal Almanac 1997. Kansas City, Kansas: Andrews and McMeel.