Gps

GPS for GIS: What’s Really Involved?
Yousif Al-Ghamdi [email protected] Mark Bowhay [email protected] and Thamer Abulleif [email protected] Surveying Services Division, Saudi Aramco Dhahran, Saudi Arabia
Copyright © Saudi Aramco, 2007

ABSTRACT The growing utilization of GPS in GIS data collection requires the GIS community to have a clear understanding of GPS technology and its relationship to GIS as a common source of accurate and reliable geospatial information. Whether using a multi-thousand dollar, sophisticated GPS receiver or a low-cost, hand-held GPS receiver; the collected data can be useless without understanding the capability of the system and the collection method; the associated geodetic datum and coordinate system information; and the geodetic structure of the GIS system. Following a general introduction to GPS technology, this paper distinguishes between the main types of GPS positioning and their capabilities. A focus on geodetic datums, coordinate systems, and transformations with respect to GIS data collection using GPS will be addressed along with a discussion on managing the quality of the collected data in GIS. Key Words: GPS, GIS data collection, datums, coordinate systems, transformations.

1

1. INTRODUCTION When the first satellite was launched in 1978, the Global Positioning System (GPS) was only intended for US military applications. However, the system was to be made available for civilian uses in mid 1980s after realizing its importance in aviation safety following a civilian airliner tragedy. By 1994, a complete constellation of 24 satellites was in orbit initiating the system’s full operational capability. Simply put, GPS is a radio-navigation technology that uses satellite signals to calculate the position of objects on the earth's surface. Today, GPS has become the world’s premier position, navigation and timing information service. The system serves millions of civil users with handheld, vehicle-mounted, and airborne GPS receivers. As GPS applications have been rapidly growing, so have Geographic Information Systems (GIS). In the late 1990s, the two technologies came into convergence when GIS professionals realized the value of GPS in acquiring and/or updating GIS base map with both location and attribute information. Nonetheless, using GPS to acquire mapping location and attribute information for features in GIS is one of the biggest challenges many agencies face in creating and maintaining comprehensive databases. This is because the positional accuracy required varies significantly. For instance, in urban planning, meter accuracy is more than enough; but for maintaining municipal water systems, centimeter accuracy is necessary. The time has arrived for the GIS community to fully understand what used to be done for centuries by ‘surveyors only’, in terms of simple field data collection. To do that, however, three prerequisites are essential for any GIS professional. These are: understanding GPS technology, understanding the geodetic aspects of GPS, and knowing how to properly use GPS for GIS. The objective of this paper is to share with the Saudi GIS community the basics of GPS and its geodetic aspects in order to help improve the Kingdom’s Spatial Data Infrastructure in terms of accurate and reliable GIS information. 2. GPS TECHNOLOGY The utilization of GPS technology for GIS data collection has notably increased in the last few years mainly due to the simplicity of using this technology with advanced field data collectors and software. However, behind the apparent simplicity of GPS is an immense yet interesting assembly of today's technologies.

2

This section highlights some of the theories of GPS technology and its utilization in navigation and positioning. 2.1. What Is GPS? GPS is a worldwide radio navigation system, owned and controlled by the US Department of Defense, designed to consist of 24 satellites, their ground stations and users’ receivers which can provide the ability to accurately determine a unique address (coordinate) for each location on the earth, at any time and weather condition. 2.1.1 System Segmentation GPS is composed of three main segments as illustrated in Figure (1).

Figure (1): GPS segments [from www.garmin.com] The Space Segment is the heart of GPS which consists of at least 24 GPS satellites orbiting the earth at an altitude of about 20,000 km in 6 orbital planes with 4 satellites in each plane spaced at a separation of 60 degrees to the equator. The satellite orbital parameters are designed to provide a constellation of at least 6 satellites within line of sight of a user from almost anywhere on earth. See Figure (2).

Figure (2): GPS space segment [from www.garmin.com]

3

The Control Segment is used to track the flight path and health of the satellites by the US Military through a globally distributed monitoring station network. The tracking information is sent to a master control station which communicates with each satellite to provide navigational updates. The User Segment basically consists of the GPS users’ receivers. 2.1.2. How GPS Works GPS positioning is based on the concept of trilateration which utilizes a basic geometric principle that allows you to compute a location if you know the distances to it from other, already known, locations. Currently, GPS satellites broadcast 3 different types of data superimposed on top of the 2 main carrier frequencies forming the navigational codes/signals, see Table (1). The first data type is the almanac which contains the coarse time and status information of the satellite. The second is the ephemeris that contains the orbital information. And the third is the pseudorandom code, a complex digital code providing the satellite clock information in 2 components: the Coarse Acquisition (C/A) code intended for civilian use and the Precise (P) code intended only for the US military. The (P) code is excluded from civilian access by encryption and is thus referred to as the (Y) code.
Carrier L1 L2 Frequency 1575.42 MHz 1227.60 MHz Code\Signal C\A & P(Y) P(Y)

Table (1): GPS frequencies and their codes. GPS satellites are equipped with atomic clocks precise to within a billionth of a second. Using this precise time and by analyzing the signal, the receiver can compute the time taken for the signal to reach the receiver. This time is then transformed by the receiver to a range (distance) using the signal’s speed of light constant value of approximately 3x108 m/s. In order to use the trilateration formula, a GPS receiver must be locked on to the signals of at least 3 satellites to determine the user's 3D position. Thus, conceptually, 3 satellite ranges form 3 virtual spheres with the 3 satellites at their centers and the GPS receiver’s logical position at one of the 2 intersections. See Figure (3). However, a fourth satellite is also required to synchronize the clocks of the other three. In summary, for a GPS receiver to find your location, it has to determine two things:

4

• •

The precise location of at least four visible satellites. The distance between you and three of these satellites.

Figure (3): GPS trilateration concept [from www.trimble.com] 2.1.3. Sources of Errors There are a number of factors that can degrade GPS positional accuracy. The first to consider is the Satellite Geometry, more generally referred to as the PDOP (Positional Dilution of Precision), which is essentially a value expressing the quality of the geometry of satellites with respect to each other from a receiver's point of view. Ideal satellite geometry exists when the satellites are located at wide angles relative to each other. Poor geometry results when the satellites are located in a line or in a tight grouping. The higher the PDOP value, the less accurate the position solution. PDOP values below 4.0 are generally considered good for accurate data collection. It should be noted that the more the satellites, the better the solution redundancy, and thus the greater the accuracy. Another consideration is the accuracy of the Ephemeris or Satellite Orbital Parameters which are continuously affected by solar wind and gravitational effects of the sun, moon and other celestial bodies. While these are continuously measured and updated, there is still a low but significant effect on real time positional accuracy. The Multipath effect describes the reflection of satellite signals from terrestrial objects such as buildings. This increases the travel time of the signal, thus causes an apparent increase in distance from the satellite. While this effect can be reduced by advanced antenna design and signal processing, it is still important to avoid data collection close to high buildings for example.

5

Figure (4): The GPS multipath error [from www.wikipedia.org] Satellite Clock errors, while small, still have an appreciable effect on accuracy. Even though the observed satellites and receiver are effectively synchronized in time, there are still signal processing and calculation errors that reduce positional accuracy. Atmospheric Effects provide one of the major error sources due to the reduced speed of the transmitted satellite signals when entering the ionosphere and troposphere. Advanced GPS receivers use built-in models that calculate an average amount of delay to partially correct for this type of error. In summary, an indication of the relative scale of the errors in a typical navigational receiver without augmentation can be seen in Table (2). Note that the errors are not cumulative.
Source of Error Ephemeris Multipath Satellite clock Tropospheric effects Ionospheric effects Accuracy Effect ± 2.5 meter ± 1 meter ± 2 meter ± 0.5 meter ± 5 meter

Table (2): Summary of GPS sources of errors [from www.wikipedia.org] 2.2. GPS Receiver Types A GPS receiver is a special radio receiver used to detect and decode GPS signals and return the output to the user in a useable form. It is composed of an antenna, tuned to the satellite frequencies, a processor and in most consumer units a keypad and display to show navigational information to the operator. The characteristics of the receiver depend on the application for which it was

6

designed. Nonetheless, the number of frequencies that the receiver can track and process is what is commonly used as a basis of GPS receiver classification. Along the same line, there are currently 2 main types of GPS receivers used in GIS data collection. 2.2.1. Single Frequency Receivers (L1) The consumer versions of these receivers determine position by processing information found in the C/A code that is transmitted by the satellites on the L1 carrier only. The main advantage of this type is the low cost of production. The main drawback is that only moderate accuracy on the order of 10-15 meters is achievable without augmentation. While early receivers used multiplexed channels to save processor costs, most recent receivers use at least 12 channels, one channel for each satellite signal. 2.2.2. Dual Frequency Receivers (L1/L2) These receivers are more sophisticated as they are capable of tracking both L1 and L2 frequencies. They are capable of determining positions by processing measurements of the carrier phase of the satellite signals over time. They do not need to decode the information being transmitted except for locating the satellites. These receivers are traditionally used for surveying and geodesy due to their high accuracy. Although they are not normally used for GIS data collection, better atmospheric modeling calculations and the enhanced processing capabilities due to signal redundancy mean that these receivers are becoming applicable for global decimeter positioning services (discussed in section 2.3) that are increasingly used for accurate GIS data collection. 2.3. Positioning Methods and Accuracies Positioning methods and accuracies can basically be summed as falling under the following categories: Autonomous Navigation, Augmented Navigation and Positioning (generally referred to as Differential GPS – DGPS), and Real Time Kinematic (RTK) Positioning. 2.3.1. Autonomous Navigation Autonomous navigation is the basic method of GPS positioning where no processing aids (augmentation) are provided to the GPS receiver. This method is the most commonly used and provides the basic capability of any type of GPS

7

receiver. Single frequency (L1) hand-held or vehicle-mounted receivers are common autonomous navigators that achieve 10–15 meters accuracy. 2.3.2. Augmented Navigation and Positioning DGPS is used to improve the accuracy obtained from GPS. Corrections are calculated at one or more locations (base stations) by comparing the known position at the base station with the satellite data computed position. From this comparison, corrections can be computed which may be in terms of position correction but are more often expressed in terms of satellite receiver distance (pseudorange). Corrections are transmitted to the user’s receiver (the rover) by various means including telephone links, radio telemetry or through satellite links. See Figure (5). To reduce costs, satellite links usually use L-band frequencies, enabling the GPS receiver to use one of the standard channels to process the corrections. Several types of GPS positioning methods fall under this category as follows:

Figure (5): Differential GPS1 2.3.2.1 DGPS Code Phase Positioning Also known as pseudorange differential, DGPS code phase positioning refers to relative positioning using the C/A code-phase observables providing meter level precisions. Although it has limited application to detailed engineering control surveying and topographic site plan mapping applications, DGPS code-phase positioning is widely used for general reconnaissance surveys, navigation,

1

Courtesy of Peter H. Dana, The Geographer's Craft Project, Department of Geography, The University of Colorado at Boulder.

8

preliminary stake out surveys, and hydrographic surveys. Terrestrial worldwide radio beacons transmitting DGPS radio signals fall under this category. Sub meter DGPS positioning can be achieved through simple single frequency hand-held GPS and DGPS receivers up to hundreds of kilometers from a base station and over continental distances using multi base station network. The accuracy of DGPS generally degrades with increased distance from a base station. This is mainly due to the influence of the atmospheric effects. Dual frequency receivers allow for more accurate positioning by providing methods for the computation of the atmospheric corrections at the user's receiver. 2.3.2.2 Satellite Based Augmentation Systems (SBAS) SBAS is the generic term used for a number of augmentation systems (DGPS) introduced in a number of areas. It is critically important for GPS users to confirm that any SBAS signal received by their GPS receiver is applicable to their location, as the broadcast signal coverage usually exceeds the area covered by the correction algorithms. WAAS (Wide Area Augmented System) is an accurate SBAS navigation system initially developed for civil aviation by the United States Department of Transportation. The system augments GPS to provide additional accuracy and availability necessary to enable users to rely on GPS for accurate positioning. Designed to provide an accuracy of 7.6 meters or better (both vertically and horizontally), actual performance (excluding receiver errors) has been measured as being better than 1 meter horizontally and 1.5 meters vertically in the United States. EGNOS (European Geostationary Navigation Overlay Service) is an SBAS under development by the European Space Agency to provide a similar service to Europe as is provided to North America by WAAS. While still under development, initial operations started in July 2005 and is due to be certified in 2008. Even though the EGNOS signal is receivable in the Middle East, the network of base stations does not cover the area and the service should not be relied on, and may in fact adversely affect autonomous positioning. 2.3.2.3 OmniSTAR DGPS OmniSTAR is typical of the commercial fee-for-service wide area DGPS systems using geo-stationary satellite broadcast techniques to deliver accurate GPS corrections. Data from many widely spaced reference stations is used in a proprietary multi-site solution to achieve sub-meter positioning over most land

9

areas worldwide. OmniSTAR is a proprietary system operated by the FUGRO group. The OmniSTAR concept is illustrated in Figure (6).

Figure (6): 1. GPS satellites. 2. Multiple OmniSTAR reference stations. 3. GPS corrections. 4. GPS monitor network control centers, corrections checked and repackaged for uplink 5. Geo-stationary satellite. 6. Satellite broadcasts footprint on earth (OmniSTAR user area). 7. Correction data received/applied real-time.1 2.3.2.4 OmniSTAR HP The OmniSTAR-HP (High Performance) solution has revolutionized the use of DGPS in precision mapping applications over entire continents without the need to deploy local base stations. It is a dual frequency GPS augmentation service that uses differential carrier-phase positioning with multiple reference stations and eliminates modeled errors. This makes OmniSTAR HP capable of providing positional accuracy of better than 10 cm (95%) horizontally and better than 20 cm (95%) vertically up to 1000 km away from reference points. The OmniSTAR HP (1000 km) coverage map for Saudi Arabia is shown in Figure (7).

1

Courtesy of OmniSTAR USA, Inc.

10

Figure (7): OmniSTAR HP service coverage map showing 1000 km radii from the four available reference points in the Middle East.1 2.3.2.5 OmniSTAR XP and StarFire Omnistar XP and StarFire are both wide area differential systems covering the entire planet, both use dual frequency GPS receivers and signal processing to compute atmospheric corrections. This, together with orbital and clock correctional information broadcast by L-Band satellite based transmitters, enable decimeter level accuracy worldwide. The Omnistar HP system automatically transfers processing to the Omnistar XP correction service when the base station correction degrades to less than HP levels of accuracy. 2.3.3. Real Time Kinematic Positioning Although technically falling under DGPS as a Carrier Phase Positioning technique, RTK is considered as another GPS positioning method due to its uniqueness in requiring 2 sets of GPS receivers within a few kilometers. RTK is used to obtain the highest precision from GPS and has direct application to most surveying and engineering mapping applications providing centimeter level accuracy. This method measures a 3-D baseline vector between the base station (receiver occupying a known point) and a second receiver at another point (rover) resulting in a vector difference between the two points occupied. Drawbacks to RTK systems include the requirement of two sets of expensive GPS receivers, radio links, at least a 2-man and 2-vehicle crew, and the limitation of a ±10 km base-to-rover distance. 3. GPS GEODETIC ASPECTS One of the least understood concepts of mapping is the role of geodesy in spatial data representation. The underlying principle of using GPS for positional data
1

Courtesy of FUGRO.

11

collection depends on geodesy. According to ESRI’s Dictionary of GIS Terminology, Geodesy is the science that determines the size and shape of the earth and measures its gravitational and magnetic fields. GIS professionals are not expected to study geodesy, but they should have a good knowledge of its basics: coordinate systems, datums, and transformations. It is arguably correct that these are the most difficult technical issues facing GIS professionals worldwide. Datums and coordinate systems are important because they form the basis from which features on the earth’s surface will be depicted. They work together to provide the locational tools needed to depict the features and phenomena of interest in GIS analysis. 3.1. Coordinate Systems In order to correctly represent the geographic location of features on a map, a mathematical representation system (coordinate system) needs to be defined so that features are located on the map relative to their location on the ground. 3.1.1 Local Coordinate Systems A local coordinate system, also known as an assumed coordinate system, is generally a plane XYZ coordinate system with an assumed value reference point. It is used particularly for project locations, sites, or plants to simplify the design and construction layout. Unfortunately, the simplicity and practicality of local coordinate systems can only be applied to a limited area. This is simply because of the earth is not flat. 3.1.2 Geographic Coordinate Systems To facilitate the correct global mathematical representation of the earth on a map i.e. taking into account the curvature of the earth, there needs to be a coordinate system that best describes the shape of the earth. Geodesists approximate the shape of the earth to be best described as an ellipsoid because its polar diameter is about 0.33% shorter than the equatorial diameter. The ellipsoid is created by rotating an ellipse around its semi-minor axis (line from the North to South poles, also known as the rotational axis). It is defined by the semi-major axis (a) and semi-minor axis (b) or flattening (f) which is a measure of how much an ellipsoid differs from a sphere, computed as:
f = a−b a

(3.1)

12

Geographic (also called geodetic) coordinates are defined by latitude, longitude, and height above the ellipsoid (ellipsoidal height). In Figure (8), the line going through point P is called the normal to the ellipsoid. It is perpendicular to the ellipsoid at that point and it is used to define the latitude, longitude, and ellipsoidal height of point P as illustrated.

Figure (8): Ellipsoid and geographic coordinate system definition.1 Another representation of geodetic coordinates is through the Cartesian coordinate system where XYZ coordinates originating at the center of the ellipsoid (earth-centered) are used. In this regard, the Z-axis is aligned with the ellipsoid’s semi-minor axis; the X-axis is in the equatorial plane passing through the 0° longitude (Greenwich meridian); and the Y-axis is in the equatorial plane passing through the +90° longitude. As the ellipsoid best represents the earth mathematically, a better shape of the earth in terms of its gravity is the geoid. The geoid represents an equipotential surface of the earth that is everywhere perpendicular to the direction of gravity. As a good approximation, the geoid is often referred to as Mean Sea Level (MSL). While this is not theoretically correct, the deviation is only about 1 meter worldwide. Since the mass distribution of the earth is not uniform and the direction of gravity changes accordingly, the resultant shape of the geoid is irregular.

1

Courtesy of Peter H. Dana, The Geographer's Craft Project, Department of Geography, The University of Colorado at Boulder.

13

The separation between the geoid and the ellipsoid is called the geoidal undulation or the geoid-ellipsoid separation. Figure (9), describes the relationship between the geoid and the ellipsoid. The formula used to derive the geoidal undulation at a point is:
H = h−N

(3.2)

Where: (H) is height above geoid or Mean Sea Level height. (h) is ellipsoidal height. (N) is geoidal undulation.

Figure (9): Relationship between the geoid and the ellipsoid.1 3.1.3 Projected Coordinate Systems A simple way of mapping the earth with minimal distortion is to map it on a globe. However, globes are expensive, cumbersome, awkward to store, difficult to measure and draw on, and less than half is visible at one time. To eliminate such disadvantages, maps are drawn on flat surfaces, or planes. It is impossible to project the surface of a spheroid exactly onto a plane. Consequently, whenever a large area of the earth is represented on a plane some significant distortion always occurs. The transfer of a spherical or ellipsoidal surface onto a flat surface is called ‘map projection’. Projections represent the geographic grid as arranged on a twodimensional surface. This arrangement is designed to best facilitate the objective of the map to be produced. Several methods have been used to project the earth’s surface into a plane. However, in practice, Universal Transverse Mercator (UTM) and Lambert Conformal Conic (LCC) projections account for 90% of base maps worldwide (J. Illife, 2000).
1

Courtesy of Smith, J.R. Introduction to Geodesy: The History and Concepts of Modern Geodesy. New York: John Wiley & Sons, Inc. 1997.

14

Figure (10): UTM zones in Saudi Arabia. Transverse Mercator projections result from projecting the sphere onto a cylinder tangent to a central meridian. Rotating the cylinder every 6° of longitude creates the UTM grid projection. Accordingly, the world is divided into 60 north-south zones, each covering a strip 6° wide in longitude. In Figure (10), UTM zones 37 to 40 cover Saudi Arabia while zone 39 is taken as an example to define its XY coordinate system. Each zone has its center at the intersection of the zone’s central meridian and the equator, corresponding to 500,000m Easting (referred to as False Easting since the value was assumed) and 0m Northing. The Lambert conformal conic (LCC) projection results from projecting the sphere onto a cone tangent to two lines of latitude (standard parallels). Only the two standard parallels are theoretically distortion-free. In theory, while UTM projection can allow minimum distortion in an area that is predominately in north-south in extent, LCC does the same for areas that are predominately in east-west extent. Hence, LCC has been used to represent the Saudi Arabian base map with two fixed standard parallels, 21° N and 27° N. 3.2. Geodetic Datums A datum can simply be defined as a reference to which a coordinate system is related. Datums can be either horizontal or vertical, or sometimes both. Using house construction as an example, when setting windows at a certain height above the floor, the floor becomes the vertical datum. Also, the horizontal and

15

vertical datum for the house can be defined as the parcel; and where the house is placed on the parcel is accomplished by using a coordinate system. 3.2.1 Horizontal Datums In surveying and geodesy, a geodetic datum is a reference point on the earth's surface against which position measurements are made, and an associated model of the shape of the earth for computing positions. Horizontal datums are used for describing points on the earth's surface, in latitude and longitude or other coordinate systems. See Figure (11) for more details.

Figure (11): Datums and coordinate systems [from www.navigator.navy.mil] There are hundreds of locally-developed horizontal datums around the world, usually referenced to some convenient local reference point. They are based on different predefined ellipsoids, i.e. coordinate systems, which best approximate the shape of the earth. The dimension and orientation of the ellipsoid are chosen "best fit" to the geoid over the specific area intended for the datum providing minimal geoidal undulation. The World Geodetic System of 1984, WGS-84, is the datum used by the US Military for GPS. It is a global best-fit datum that uses the WGS-84 ellipsoid i.e. observed GPS coordinates are latitudes, longitudes and ellipsoidal heights based on the WGS-84 datum. Furthermore, a global network of accurately positioned ground stations comprises another global datum, called the International Terrestrial Reference Frame (ITRF). The ITRF is accurately positioned relative to the geocenter using a variety of space-based techniques. It has been created in a number of time-based realizations since 1989 with the latest being ITRF-2000 which is a more recent datum than the WGS-84 datum and is considered to be an improvement upon WGS-84 with increasing international recognition. Ain Al Abd-1970 is considered Saudi Arabia’s local horizontal datum and is being used by several mapping and GIS agencies in the Kingdom. A list of common

16

reference ellipsoids used by local, regional, and global datums is provided in Table (3).
Datum Ain Al Abd-1970 North American-83 ITRF-2000 WGS-84 Ellipsoid International-1924 GRS-80 GRS-80 WGS-84 a (meter) 6 378 388 6 378 099 6 378 137 6 378 137 1/flattening 297 298.257222101 298.257222101 298.257223563 Location Saudi Arabia North America Global Global (GPS)

Table (3): Common datums and their reference ellipsoids. 3.2.2. Vertical Datums Vertical datums are used for measuring the elevations of points on the surface of the earth. They are either tidal, based on sea levels, providing heights referenced to sea levels; or geodetic, based on the same ellipsoid used for the horizontal datum, providing heights above the ellipsoid. Commonly, elevations are given in heights above ‘Mean Sea Level’ (MSL) so the average of tide heights over a period of many years is used because sea tides cause water levels to change constantly. 3.3. Geodetic Datum Transformations The coordinates of a point will change depending on which datum the coordinates are referred to. To change a coordinate from one datum to another, a mathematical process known as transformation is used which requires a number of points with positions known in terms of both datums i.e. common points. The accuracy of the transformation depends on the method chosen, and the accuracy, number, and distortion of the common points. Geodetic datum transformations are usually geocentric 3 or 7 parameter, or gridbased. Local transformation methods are outside the scope of this discussion as the focus is on earth-centered coordinate systems only. 3.3.1. High Accuracy Transformation A grid-based transformation model is considered the most accurate, consistent, and efficient transformation suitable for precision mapping over large areas such as Saudi Arabia. This kind of transformation became important particularly after the development of high precision GPS, such as RTK, in order to maintain the same relationship over local datums. Basically, Grid-based transformation models can be either horizontal (transforming between the latitudes and

17

longitudes of 2 datums) or vertical/geoid (transforming between ellipsoidal heights and MSL elevations). 3.3.1.1. Horizontal Grid-based Transformation For seamless transformation methodology over large areas between latitudes and longitudes of 2 different datums, a horizontal grid-based transformation model is the most appropriate. Generally, the model consists of a pair of binary (.las and .los) files that contain coordinate shifts for latitudes and longitudes, respectively, obtained from analyzing 2 sets of common points on the 2 different datums. The technique is to use the difference in position at each common point to in effect fit a surface, which can then be used to interpolate points on a grid to model the change for other points. In practice, a bi-linear interpolation process uses the model to easily, on-the-fly provide, for instance, Ain Al Abd-1970 coordinates from WGS-84 GPS coordinates. 3.3.1.2. Geoid Grid-based Transformation Similarly, a geoid grid-base transformation model is best utilized for wide-area transformation of ellipsoidal heights to Mean Sea Level (MSL) elevations. Commonly, the model is a .bin format file composed of geoidal undulation (N) values representing the separation between an ellipsoid and a geoid (MSL) based on Equation (3.2). In practice, a bi-quadratic interpolation process using the model is performed to provide on-the-fly MSL elevations from GPS ellipsoidal heights. 3.3.2. Medium Accuracy Transformation The relationship between two consistent, 3-D coordinate systems can be defined by a 7-parameter transformation. The 7-parameter method assumes a similarity relationship between the 2 datums, a local and a global datum for instance. Figure (12) illustrates these parameters which are listed here: A. 3 origin shifts (∆X, ∆Y, ∆Z) B. 3 rotations around the X, Y, and Z axes (ε, ψ, and ω respectively) C. 1 scale change (S) A minimum of 3 common points are used to solve for 7 parameters which represent the relationship between the two datums. In fact, many more are needed to reliably model the relationship. If used over a large area, Saudi Arabia

18

for instance, a 7-parameter transformation could be suitable for medium accuracy projects, of the order of 1 m. However, a 7-parameter transformation can be sufficient for precision mapping for small areas, a city for example. It may be possible to produce separate 7-parameter sets for sub-regions to enable consistency in transformation over the specified areas.

Z’ Z

Datum # 1
X’

ε
∆ ∆

ω ψ
Y
∆

Datum # 2
Y’

X

Figure (12): 7-parameter transformation [from www.navigator.navy.mil] 3.3.3. Low Accuracy Transformation A 3-parameter transformation, such as Molodensky’s transformation method, is considered a low accuracy transformation which uses the 3 origin shifts only (∆X, ∆Y, ∆Z), see Equation (3.3). When transforming between 2 datums of different ellipsoids, the difference between the 2 ellipsoids, local and global for example, in terms of semi-major axis and flattening is considered. The origin shifts can be determined by an averaging of the same differences at each of the common points and the difference in ellipsoids is a simple subtraction of the ellipsoid parameters.
X’ Y’ Z’ X Y Z ∆X + ∆Y ∆Z

=

(3.3)

This method of transformation is useful only for navigation purposes outside very limited areas and so is generally not applicable to precision mapping. It is often used in hand-held GPS receivers with the 3-parameters published by the United States National Imagery and Mapping Agency (NIMA) to transform between many local datums and WGS-84. Table (4) contains these parameters, along with their error estimate, in meters and the difference in ellipsoid values for Ain Al

19

Abd 1970 transformation to WGS-84. As you can see, the accuracy that could be expected from this method would be about 10 meters.
Datum Ain Al Abd 1970 Ellipsoid International 1924 ∆a (m) -251 ∆f x 10
4

∆X (m) -143 ±10

∆Y (m) -236 ±10

∆Z (m) +7 ±10

-0.14192702

Table (4): 3-parameter transformation plus datum shifts to WGS-84 4. GPS FOR GIS 4.1. Managing GPS Data in GIS There are preliminary issues as well as data quality management aspects that GIS professionals should consider to maximize the value the collected GPS data. 4.1.1. Preliminary Issues Initially, it is essential to identify the accuracy requirements of the targeted GIS application. Also, the user should take into consideration the future value of acquiring accurate data from the outset. This is done by identifying the most costeffective GPS acquisition method available for the user’s GIS application accuracy requirements i.e. RTK, DGPS, etc. The equipment costs are almost certainly small in relation to the long term value of the collected data. However, the operating costs are high. Therefore, the user ought to maximize the value of field acquired data by insuring that the positional accuracy and data quality will fulfill all future GIS and mapping requirements. Good GIS data model design is imperative. It can be difficult and expensive to expand a GIS design to include additional requirements such as positional metadata (location, datum, method and accuracy) of say, point data. Therefore, the user should make provision for future requirements from the outset. For example, it is important to clearly differentiate between ‘land form’ (topography) and ‘land use’, land use can and will overlap land form. Model design affects fundamental software issues such as the implementation of scale related viewing which has a major effect on the usability of data on mobile viewing platforms. 4.1.2. Data Quality Management GIS users need to be cautious when managing GIS data. Although there are many sources of physical, economic and demographic GIS data, importing and combining this data in a GIS is not always easy or inexpensive. This is because data often comes in different datums/coordinate systems or because it

20

sometimes lacks metadata (i.e., documentation of what the data is, data source, geodetic information, definitions of the attribute data variables, etc.) Users need to gather as much documentation about the data as possible. It is essential to appropriately track the input of GIS data by identifying the source and accuracy of all data. To borrow a quote from philosophy, ‘those who forget their history are condemned to repeat it’, is as applicable to GIS as it is to philosophy in the means of GIS metadata. Another important aspect of GIS data quality management is the representation of the actual location of points (surveyed positions). These should be stored in GIS as attributes because topology requirements and the use of GIS data of varying accuracies may require that the displayed position of a point feature may not be coincident with the surveyed location. For example, the precise GPS survey locations of traffic lights could show a lateral displacement when plotted on a dataset created from less precise aerial or satellite based photogrammetry as seen when comparing the 2 parts of Figure (13). For topological reasons, the plotted location of the traffic lights should be adjusted to fit the road data. However, for quality checking and future control reasons, the surveyed locations should be stored in the data base as attributes.

Figure (13): Lateral Displacement of traffic lights in a GIS graphical view. As another example, a pipeline has been acquired from small scale photography with an estimated accuracy of 3m while valve locations have been surveyed by RTK GPS with estimated 10cm accuracy. The valves will not, most probably, coincide with the pipeline and will need to be snapped onto it to maintain the topological relationship. For both field location purposes and future updating from a more accurate data source, the true geographic location of the valve should be stored as attributes.

21

4.2. GIS Data Collection A GIS data collection system is a combination of data collector (field computer), GPS receiver, and data collection software. These may be integrated in some cases, data collector with built-in GPS and software, but are often user assembled. 4.2.1 Data Collection Software For seamless GIS data collection using GPS, it is very important to maintain a direct relationship between field data collection/editing and the office-based GIS software. The systems adopted should interact with each other (field and office) with a minimum of operator intervention. Ensure that the field data collection software is compatible with both the officebased GIS datum/coordinate system and with the GPS datum/coordinate system. Either the data collection software or the GPS receiver must be capable of realtime transformation of the GPS coordinates to the local coordinate system framework. To achieve High Precision GIS data collection using GPS, some form of datum transformation will probably be required. GPS settings may be restricted to 3 or 7 parameter transforms, but data sets covering large areas, and international correction services may require a more complex solution. For instance, if using external sources of GPS augmentation, such as OmniSTAR’s HP, the user must make sure that the relationship between the GIS datum and the provider’s datum is known. Even though most GPS receivers and GIS office software have built-in 3-parameter datum shifts from WGS-84 to almost all local datums, selecting the appropriate datum transformation method is mandatory in order to exploit the accuracy of the augmented GPS. The data collection software should also support the data communications protocol of the GPS receiver. While the National Marine Electronics Association (NMEA) 0183, version 2.0, data format is widely supported, the standard sentence used in GIS data collection is the GGA sentence shown below:
$GPGGA,121505,4807.038,N,01131.324,E,1,08,0.9,133.4,M,46.9,M, ,*42

The default format of this sentence provides Latitude, Longitude, Mean Sea Level Elevation and Geoidal Undulation but only to about 0.75m horizontal accuracy. High precision GPS requires an extended length data format for centimeter-accuracy Latitude and Longitude.

22

While a simple vertical transformation using Equation (3.2) can be used for transforming GPS WGS-84 ellipsoidal heights to Mean Sea Level elevations over small areas, the data collection software may need to be capable of supporting Geoid Grid-Based Modeling. The default Geoid Model used by most GPS receivers, normally a global Earth Geoid Model (GEM), will not be accurate enough for high precision GIS data collection over large areas. Preferably, the data collection software will support user customization. At the very least, it should be possible to create custom forms with default values for standard attribution. Attribute fields should also support predefined tables to ensure correct values are supplied to the system. 4.2.2 Data Collectors The data collector, or the field hand-held computer, is one of the most important and problematic components of a GIS field data collection system. It should be able to withstand field operations in harsh conditions, be readable in variable lighting conditions, easy to operate, easy to carry and must be able to support the data collection software. The following are some of the specifications that are normally required: Rugged: While rugged field data collectors cost between 2 and 10 times the cost of an equivalent consumer device, they should be very seriously considered, particularly when it is anticipated that they will be used full time. Sunlight Readable: Most Pocket PC devices are readable in direct sunlight, but most Tablet PCs are not. Daylight readable does not always mean sunlight readable; unless a transmissive screen has very bright backlighting, it will probably not be readable in direct sunlight. Reflective screens use a mirroring technique to reflect incident sunlight back through the display and are almost always readable under any bright sunlight conditions. Transflective screens, combining the best of both technologies, are probably the best option. Optimal Screen Size: Basically, the bigger the screen size, the better, as long of course as the device is still portable. Standard 1/4 VGA (320x240) are the smallest size that should be considered. These are used by almost all rugged PDA devices with a very few using half or full VGA (640*480). Larger screen sizes are almost exclusively the domain of Tablet PC devices. SVGA (800x600) and XGA (1024x768) are the norm. Compatible Operating System: This will normally be decided by the choice of data collection software. A few packages, including ESRI's ArcPAD, will operate on Windows Mobile (CE) or Windows XP. Most packages however will be

23

supplied for the Windows Mobile device while a few still use proprietary systems. But the user should not assume that a later version of an operating system maintains all the capabilities of a previous i.e. test any applications before committing to a new Operating System. 4.2.3 GPS Receivers The operational accuracy requirements should be decided upon by the user depending on the type of application. They determine the choice of the GPS receiver and affect the selection of the data collector and the software. The following is a summary of operational accuracy requirements along with examples of GPS equipment selection and estimated costs: 4.2.3.1 Navigational or Recreational (±15m)

For point to point navigation without maps: Garmin: eTrex , Geko 101, GPS 76 With manufacturer or third party mapping: Garmin: GPS V, GPSMAP 76CS Estimated costs: $100 to $600 4.2.3.2 Navigational or Recreational with User Mapping (±15m)

Any navigational GPS receiver with NMEA output: $100 - $500 Garmin GPS 35, GPS 17HVS, GPS 10 ESRI ArcPAD 7 from ESRI: $500 Windows CE PDA: consumer version ($300 - $600), rugged version ($1,500 - $ 4,000) Total estimated costs: $1,000 to $5,000 4.2.3.3 GIS Attribute Collection (±5m)

Similar to Navigational (above), depending on positional accuracy required to identify feature to be attributed, a Beacon DGPS receiver may be required. ESRI ArcPAD 7 ($500) with Windows CE PDA ($300 to $4,000)

24

Any Navigational GPS Receiver with NMEA output plus DGPS Receiver DBR 300, MBX-3S, GeoBeacon ($500-$1,200) OR Trimble GPS Pathfinder Pro XRS receiver ($5,200) OR survey grade GPS receiver with Beacon DGPS and OmniSTAR Satellite based VBS DGPS Total estimated cost: $2,500 to $10,000 4.2.3.4 GIS Feature Collection for Position and Attribution (better than ±1m horizontally and ±2m vertically)

Feature collection normally requires a more accurate GPS system than those used for attribute collection. OmniSTAR 3200LR12 Receiver, GPS & Satellite based DGPS ($5,000) OR Trimble GPS Pathfinder Pro XRS receiver with OmniSTAR VBS submeter DGPS correction service ($1,800 per annum) ESRI ArcPAD 7 plus Trimble GPS Correct ($500) OR Trimble TerraSync GIS data collection software plus rugged Windows CE PDA Total estimated cost: $10,000 plus $1,800 service charge per annum. 4.2.3.5 Mapping and GIS Data Collection (±10cm horizontally and ±20cm vertically)

Mapping applications need a GPS receiver and positioning service at least as accurate as the mapping and preferably an order of magnitude better to provide data and control. OmniSTAR 8300HP dual frequency GPS and HP differential correction receiver ($7,000), other manufacturers’ receivers are available. OmniSTAR HP/XP 10cm correction service ($3,500 per annum) ESRI ArcPAD 7 plus rugged Windows CE/Mobile PDA. Total estimated cost: $10,000 - $12,000 plus $3,500 annual service charge. 4.2.3.6 Engineering and Surveying (better than ±5cm horizontally and ±5cm vertically)

The price of an Engineering and Surveying RTK GPS system is given here as an indication of the escalating costs of increasing accuracy. Permanent base

25

station systems can be used as an alternative to the two set RTK system but may not cover all areas of operation and will entail a service fee. Two dual frequency Survey Grade GPS Receivers (Trimble 5700/R7/R8) plus or including an antenna in an RTK Package ($ 35,000 to $50,000) Trimble TSC2 with Survey Controller Software ($5,500), Office Software and accessories (Trimble Geomatics Office) normally are part of the package. Radio Modem Communications link, Pacific Crest PDL ($3,500), a radio license is required. Typical system estimated cost: $40,000 - $60,000. 4.2.3.7 Survey Control (±2cm horizontally and ±2cm vertically)

The cost effective establishment of survey control is even more expensive involving a minimum of two but preferably three dual frequency survey grade GPS receivers. Trimble 5700/R7 plus Antenna ($16,000 - $25,000 each set) Survey Controller or Laptop computer with software ($5,500) with Office Software including baseline and network processors ($5,500) but may be included if purchased as a package. Typical system estimated cost: $45,000 - $75,000. 4.2.4 GPS Synopsis A field GIS data collection system based around a local, national or internationally based 10cm augmentation system will normally give sufficient accuracy for the location of all GIS features, including buried items, as well as data for all but detailed engineering design. While many relatively densely populated countries have established national RTK compatible (2-3 cm) augmentation services, it is unlikely that such systems will ever be economic in Saudi Arabia. Major conurbations in the Kingdom may well put these systems into operation but to achieve nationally, the accuracies advocated in this paper, the only current economic solution is the commercial 10cm global service offered by OmniSTAR and StarFire.

26

5. CONCLUSION Cost-effective, practical, and attractive GPS technology is used by an increasing number of GIS professionals and users for today’s GIS data collection. On the other hand, GPS is not a closed-box technology and should be utilized with a reasonable sense of theoretical and practical knowledge. This paper discusses, on a first-things-first basis, what GIS professionals and users should know before starting GIS data collection using GPS. Only after understanding the theories of GPS technology, can we relate it to geodesy in terms of coordinate systems, datums, and transformations. These concepts are essential to effective, reliable, and accurate GIS data collection. The last part of the paper discusses preliminary issues related to using GPS for GIS data collection, defines a field GIS data collection system, and provides information on appropriate system combinations for every operational accuracy requirement.

27

REFERENCES Al-Ghamdi, Y. and M. Bowhay. “Assessment of GIS Data Collection Software for High Precision DGPS in Surveying and Engineering Mapping Applications.” The First National GIS Symposium in Saudi Arabia. Khobar, Saudi Arabia. November 21-23, 2005. Anderson, J. M. and E. M. Mikhail. Surveying Theory and Practice. Boston: McGraw-Hill, 1998. Burtch, Robert. “Datums: How to Work with Them.” IMAGIN Conference, Traverse City, Michigan, USA. 2002. Dana, P.H. The Geographer's Craft Project. Department of Geography, The University of Colorado at Boulder, USA. (September 10, 2005) http://www.colorado.edu/geography/gcraft/notes/gps/gps_f.html Dawson, J. and J. Steed. “International Terrestrial Reference Frame (ITRF) to GDA94 Coordinate Transformations.” Geoscience Australia. Version 01.03.2004. Environmental Systems Research Institute, Inc. (ESRI). Dictionary of GIS Terminology. ESRI Press. Redland, California, USA. 2001. FUGRO. Fugro SurveyPlanner. (October 4, 2005) http://www.surveyplanner.com/ Garmin International Inc. (March 3, 2007) http://www.garmin.com/ Hendrikse, J.H.M. “Use of the Spatial Reference Object Model to Enhance Projection and Datum Transformation.” ESRI User Conference – Paper 376. San Diego, USA. July 5-12, 2003. ICSM (Intergovernmental Committee on Surveying and Mapping). Geocentric Datum of Australia, Technical Manual, version 2.2. 2002. ISBN 0-9579951-0-5. Illife, J., 2000 “Datums and Map Projections for Remote Sensing, GIS and Surveying” by Whittles Publishing.

28

Datum #1

Lapucha, D., R. Barker, and H. Zwaan. “Wide Area Carrier Phase Positioning – Comparison of the Two Alternate Methods.” European Navigation Conference GNSS. Rotterdam, The Netherlands. May 2004. Leick A. GPS Satellite Surveying. 2nd Edition, 1995. John Wiley & Sons. National Imagery and Mapping Agency. “World Geodetic System 1984: Its Definition and Relationships with Local Geodetic Systems.“ Third Edition, January 3, 2000. Navigator of the Navy (February 24, 2007) https://www.navigator.navy.mil/ OmniSTAR B.V. OmniSTAR 8300HP. User Manual. Issue 1.1. June 2003. OmniSTAR USA, Inc. Worldwide Digital Global Positioning Service – Omnistar USA, Inc. – How it Works. (September 17, 2005) http://www.omnistar.com/howitworks.html Pendleton, G. “Surveying for GIS’ Sake.” American Congress on Surveying and Mapping (ACSM) Bulletin. No. 213. January/February 2005. Smith, J.R. Introduction to Geodesy: The History and Concepts of Modern Geodesy. New York: John Wiley & Sons, Inc. 1997. Trimble Navigation Ltd. (March 3, 2007) http://www.trimble.com/ US Army Corps of Engineers. Engineering and Design - NAVSTAR Global Positioning System Surveying. Washington DC. July 1, 2003. Wikipedia. (February 28, 2007) http://www.wikipedia.org/ Zhang, Larry. ‘Understanding Spatial Accuracy and Integrity in GIS.” The First National GIS Symposium in Saudi Arabia. Khobar, Saudi Arabia. November 21-23, 2005

29