2
NSDI FOUNDATION
Experience has demonstrated that collection of accurate spatial datacan be an extremely costly and time-consuming process. Strategicallyfocusing resources for data collection and maintenance efforts isa challenge to federal, state, and local governments and the privatesector. As addressed in the previous chapter, data-sharing activitiesand partnerships are becoming increasingly attractive. Both, however,require a foundation that will enable the integration of data frommultiple sources. The data that make up this foundation will createa convenient and common reference for the compilation of other spatialdata.
DEFINITION OF A FOUNDATION
A useful metaphor to understand the needs for a NSDI foundation canbe taken from building construction. A solid foundation of concreteor other material is first put in place; then a framework of steelbeams is connected to the foundation to create a structure to supportthe building’s interior and exterior. The foundation must be coherentand stable, such that the structure can be expanded upon with confidencein its reliability and integrity. In the same way, a foundation ofspatial data serves as a reference for integrating other data themes.As these themes are developed and integrated with the foundation,a structure will be created that can support and sustain the NSDI.
As used in this report, the foundation is comprised of spatial data themes that are the minimal directlyobservable or recordable data to which other data are spatially referencedand from which other digital spatial data may be compiled.
Based on its members’ collective experience, the MSC has identified three types of dataas the foundation—geodetic control, digital terrain elevation, anddigital orthorectified imagery. Geodetic control is required to systematicallyregister all other information with a locational component. Thisis a primary data set in the NSDI, as it forms the footings of thefoundation. A fully integrated NSDI cannot exist unless all datatypes are mathematically registered to a common foundation of geodeticcontrol.
Two other types of data are included in the foundation. Digital terrainelevation data add horizontal and vertical measurements to the foundationand provide a fabric that approximates the Earth’s surface. Digitalimagery that has been orthographically corrected records a pictureof the landscape. Both of these data types are registered to geodeticcontrol. Spatial data for many applications then can be reliablyintegrated with this foundation.
DATA THAT FORM THE FOUNDATION
Geodetic Control
The accepted geodetic reference system in the United States is theNorth American Datum 1983 (NAD-83) established by the National GeodeticSurvey (NGS) of the National Oceanic and Atmospheric Administration(NOAA) as the first-order horizontal reference system for the nation.Vertical control is referenced to North American Vertical Datum 1988(NAVD 88). This system is well monumented on the ground for bothhorizontal and vertical control,
and in many cases the control network has been densified by stateand local users. On a global scale, the World Geodetic System-1984(WGS-84) is used by the Department of Defense as its reference forthe Global Positioning System (GPS). The difference between NAD-83and WGS-84 is extremely small, and both can be used interchangeablyfor all but the most stringent situations. The NGS has a plan1 for an updated national spatial reference system based on GPS measurements.This new reference system would be made compatible with the existingNAD-83 and NAVD 88.
For many applications the geodetic control network may be transparent;that is, the geodetic data may never be “used” directly in an application.This transparency will work as long as the data that are used torepresent various spatial data themes are tied to the geodetic control.The accuracy of the location of a spatial feature will vary accordingto the goals of the application. In addition, existing accuracy requirementswill likely change with future improvements in data collection capabilitiesand/or user requirements. Because of the availability of GPS (particularlydifferential GPS technologies) and their decreasing costs, highlyaccurate locations of newly acquired spatial data are being tiedto geodetic control at an accelerating rate.
Digital Terrain
Digital terrain (elevation) data are used to create a digital representation(or model) of the Earth’s surface. Digital terrain data have manyvaluable uses.2 The data are required in the production of digital orthorectifiedimagery and together with the imagery can be used to create viewsof the Earth’s surface from any vantage point. Digital terrain dataare used to generate several important products and analyses, includingvolume, slope, aspect, line of sight,
and intervisibility. These products at appropriate resolution canbe used for civil engineering earth-work computations, stormwaterrun-off studies, microwave tower site selection, soil stability studies,geological studies, and many more. Digital terrain data also arecommonly used to create contour information on many maps.
There are two basic models used to represent digital terrain data:(1) digital elevation matrix (DEM) and (2) triangulated irregularnetwork (TIN); these are schematically shown in Figure 2. The DEM is a grid of elevation values at regular row and columnspacing and is generally defined with an origin, the number of rowsand columns and their spacing, and a series of elevation values.A DEM is an efficient method of storing terrain data; only elevationis needed as ground location is implied by row and column positionswithin the matrix. On the other hand, the TIN is a series of pointslinked into triangular surfaces that approximate the surface. Thespacing of points in a TIN are nonuniform, which allows points tobe located on critical terrain features. This offers the potentialfor the terrain to be more accurately modeled with a minimum numberof points. A TIN also allows for faithful representation of linearfeatures (geomorphic features), such as ridges, drains, and embankments.
The accuracy of digital terrain models depends on the source of thedata, the point density and distribution, and whether or not geomorphicdata were used in the production. Sources of digital terrain modelsvary—from highly accurate ground surveys (e.g., GPS positions) orlarge-scale photogrammetric surveys to lower-accuracy data createdby digitizing contours from paper topographic maps or from small-scale,high-altitude photogrammetric surveys. Regardless of the specificdata sources, digital terrain models need to be tied to a geodeticcontrol network.
sources of distortion are removed, and the image has the propertiesof scale and accuracy associated with a map. The image can also bederived from digital airborne or satellite sensors. They can representa range of resolutions based on the altitude and format of the originalphotography and the scanning process used to produce the digitalraster image. Application requirements and costs will ultimatelydictate the resolution of the imagery.
An example of a current data product that will meet the needs ofa large number of users—the digital orthophoto quarter-quad (DOQ)—is being produced by the USGS, the Soil Conservation Service (SCS),and the Agricultural Stabilization and Conservation Service (ASCS).As specified by the USGS, each DOQ covers a “quarterquad” or roughly a 4 × 4 mile area. A quarterquadimage consists of about 40 million pixels; each pixel representsa 1-m2 area and has a 256-level gray-scale value. The USGS/SCS/ASCS haveproduced technical specifications for DOQs, and several test datasets have been widely distributed and tested.
When displayed, a DOQ (see Figure 3) looks like an aerial photo on which one can identify such featuresas roads, houses, trees, and driveways. About 200,000 DOQ imageswould cover the conterminous United States. DOQs contain coordinateregistration points that permit measurement of the location of visibleobjects to a visual resolution of a few meters. This would allowfor addition of other information, such as transportation, to theimage base.
Expensive photogrammetric equipment and highly skilled technicianswere needed to create the rectified stereo images that have beenused for decades. These images have been the primary sources forupdating the USGS 1:24,000 quad paper maps. Advances in computertechnology now permit application specialists to access DOQs directly,putting a wide range of new users in touch with a rich and timelyform of spatial data. DOQs promise to be a
high-accuracy, low-cost resource supporting many map-making and geographicinformation systems (GIS) activities in government, academia, andindustry.
The value of digital orthorectified imagery is evident to a numberof states and local governments, which collect and use these dataas a critical part of their foundation for GIS activities. At thelocal level, this imagery may be at a much higher resolution thanthat of the DOQ program, reflecting specific local information needs.At the state level, the needs for the imagery are similar to thosefor the DOQs; however, current state imagery programs have resolutionsranging from 0.5 to 1.2 m pixels, and some states use color infraredinstead of the black and white image sources used in the DOQ program.By and large, the various state programs have specifications similarto those of the USGS/SCS/ASCS program.
Because of their use within a foundation for the integration of otherspatial data, national coverage by DOQs or other orthorectified digitalimagery programs should be assigned a high priority. The USGS/SCS/ASCSestimated the cost for one-time nationwide DOQ coverage at $180 millionspread over a five-year program. Federal-state-local-private partnershipscould lead to sharing of both the costs and the benefits and theestablishment of this important component of the NSDI foundation.The MSC recommends that current federal plans for DOQ productionbe accelerated and that nationwide coverage be achieved through partnershipswith states that plan or have similar programs that meet or exceedfederal specifications.
NOTES
1. Draft Implementation Plan for the National Spatial Reference System,1994, National Oceanic and Atmospheric Administration, 37 pp. Thisdraft implementation plan was reviewed at a forum by the NationalResearch Council’s Committee on Geodesy, which issued the report,Forum on NOAA’s National Spatial Reference System (1994), National Academy Press, Washington, DC, 66 pp.
2. Further discussion of digital elevation models and their use isgiven by P. A. Burrough in Chapter 3 (pp. 39-56) of his book Principles of GeographicalInformation Systems for Land ResourcesAssessment (1986), Clarendon Press, Oxford.