Issues regarding integration of GPS data for NRIS database generation

Introduction
For the management of National Resources, computerized digital databases are generated for different administrative units like village, district , state and for different themes like landuse, roads, canals, elevation points etc. At present, thematic maps generated by different agencies is the main source of inputs for the creation of National Resources Information System (NRIS) district databases. Dependency on GPS data and satellites data including stereo pair is much less. It is required to update, validate and enhance databases periodically for reliability of GIS outputs and integrity of data analysis. To meet such requirements GPS technology provides facility for data collection pertaining to new developments and also for checking the accuracy of existing databases.

After removal of Selective Availability (S/A), the applicability of GPS data in facility mapping, infrastructure planning and GIS database generation will increase by many folds. In context of NRIS, it is necessary to prepare fresh base maps for many districts using IRS data and control points provided by GPS observations due to restriction policy of digitization of SOI maps and non availability of current maps. Thus, GPS is an important source for NRIS database generation.

This paper discusses interface issues and data integration problems in context of authors experience regarding usage of GPS receivers for NRIS database validation and enhancement

GPS System
The NAVSTAR GPS (Navigation Satellite Timing and Ranging Global Positioning System ) is a satellite based radio navigation system providing precise three dimensional position , navigation and timing information to suitably equipped users on a continuous basis. GPS receivers measures code or carrier phase or both to provide meter level accuracy in point positioning mode and up to few centimeters in differential mode. The brief description of GPS system components viz. space segment, control segment and user segment is given below.

Space Segment
The space segment consists of 24 satellites arranged in six different orbital planes of inclination 55 Deg. w.r.t. equatorial plane. These satellites are orbiting the earth at a height of 20200 km from the surface of the earth and have periodic time of 12 hours. Minimum 4 satellites are visible for positioning on ground/sea/air at any time throughout the year. Each satellite is transmitting coded signals known as pseudo random noise (PRN) signals modulated on L1 (154 * 10.23= 1575.42 MHz (= 19.05 cm)) and L2 (120 * 10.23 = 1227.60 MHz (=24.45 cm)) carrier frequencies. Transmitted signals on both frequencies are modulated with navigation and system data including satellite ephemeris, atmospheric propagation correction data and satellite clock bias information. The L1 signal contains both, P- code and C/A code. The L2 signal contains P code only. The P code frequency is 10.23 MHz which corresponds to wavelength of 29.31 meter with period of 267 days and 7 days for a satellite. The C/A code frequency is 1.023 MHz which corresponds to wavelength of 293.1 meter with period of 1 millisecond. Data signal frequency is 50 bps and its cycle length is 30 second.

Control Segment
The Operational Control Segment for GPS consists of the Master Control Station near Colorado Spring (USA), three Monitor Stations and ground antennas in Kwajalein, Ascension and Diego Garcia, as well as two more monitor stations in Colorado Spring and Hawaii. The tasks of the control segment are to:

monitor and control the satellite system continuously for uploading data into the satellites.
predict the satellite ephemerides and the behavior of the satellite clocks.
periodically update the navigation message for each particular satellite.

User Segment
This relates to various types of GPS receivers like navigation, survey, single frequency, dual frequency etc.

Objectives
To study the role of GPS data in NRIS database content validation and enhancement with respect to following objectives.

Determine location accuracy and alignment of NRIS database elements like roads, railway lines, canals and selected area features like water bodies ,forest boundaries etc.
Determine height accuracy of elevation points and contours in databases
Provide accurate control points for registration and rectification of satellite images
Provide accurate control points for registration of cadastral map
Provide control points for DEM generation and evaluation
Error modeling of GPS observations.
Provide control points and Spatial Framework for base map generation

Experiment Conducted
GPS satellites are available for all 24 hours and observation can be taken at any place at any time. For better accuracy planning is required with respect to identifying area of interest, distribution of points, data collection strategy, availability of required no of satellites with good GDOP value and above 15 degree of elevation angle. Detailed procedure followed during experiment is as below,

Preparation of map for different coverages for which data are to be collected
Selection and marking of features/control points on map with identification code
Total time of data collection and assigning time slot for different area
Preparation of satellite visibility chart
Logistics planning
Selection of reference points position and rover points positions for data collection
Determining initial approximate coordinates of observation points in WGS 84 system

GPS observations were taken at well distributed points within map sheet no 46 F/9 using two LEICA SR 9400 GPS receivers one working as reference receiver and other working as rover receiver. During first schedule simultaneous observations were taken at reference point (continuously for 6 hours) and at seven rover points(for 30-40 minutes). Similar observation pattern was followed for second schedule.

Data Processing
Specific mission parameters like ambiguity resolution limit, models for atmospheric and ionospheric correction, baseline length limit etc. were defined before data processing. The coordinate solutions were determined using phase and code measurement.

The computed WGS 84 coordinates were converted to Everest Datum using computed local seven parameters. Heights above ellipsoid were converted to MSL value using external geoid model.

Results

For all observed points WGS 84 coordinates are given in table-1.
Differences in NRIS database coordinates and coordinates computed using local 7 parameters are given in table-2 for selected 8 points with larger base-length. Differences are within 10 -15 meter limits.
Differences in GPS coordinates and coordinates estimated using 7 parameters (for whole Indian region) and Geoid model are given in table-3. Comparatively larger errors are found while using global parameters for Indian region.

Conclusion
Accuracy obtained using LIECA SR 9400 single frequency GPS receiver measuring code or carrier phase is of the order of 10-15 meters in point positioning mode with 5 hours of observations. Accuracy of similar order is obtained at rover points also within distance of 2-3 kilometers for observations of 30 minutes.

Due to limitation of single frequency LIECA receiver observations for more then 45 minutes are required for base-length of 5-10 kilometers for resolving ambiguity.
The main source of error for single frequency receiver is ionospheric delay. Therefore, observations for base-length more then 10 kilometers could not be taken due to system limitation for compensating for ionospheric error
It was possible to identify and correct abnormal height observation due terrain familiarity

The desired accuracy of GPS observations can be obtained by appropriately taking care of following issues

Selection of day and time of observation
Duration of observation
Selection of reference point and corresponding rover points
Datum conversion
Height conversion

Future Issues

Develop GUI based interface for on line tracking of GPS data and converting it into ARC/INFO coverage
Create base maps using GPS observations

Acknoledgement
We are thankful to Shri A.R Dasgupta, DD SITAA for providing us opportunity to explore feasibility of using GPS data for NRIS project.

Table-1: Coordinates Derived Using GPS (WGS-84 Datum)
Point 17 is Considered as Reference Point for Baseline Processing

SR. NO.	LATITUDE	LONGITUDE	HEIGHT ABOVE ELLIPSOID
17	22 50 28.09682	73 40 24.38946	87.1877
08	22 48 08.85339	73 37 02.93546	69.3931
11	22 51 44.32849	73 37 37.03182	68.2053
13	22 54 36.81210	73 38 08.98172	65.8170
15	22 53 47.64277	73 36 01.11677	74.3100
50	22 51 44.11065	73 31 58.76126	51.1162
14	22 54 21.33026	73 37 24.27028	89.8122
16	22 52 56.72526	73 33 54.93078	71.5811
Point 3 considered as Reference Point for Baseline Processing
SR. NO.	LATITUDE	LONGITUDE	HEIGHT ABOVE ELLIPSOID
03	22 47 01.64677	73 33 49.27532	54.8687
01	22 47 44.28205	73 30 09.98229	31.9430
05	22 47 16.51705	73 41 02.23118	98.4480
06	22 48 54.08300	73 33 18.23283	54.5433
07	22 47 08.08902	73 36 40.59074	63.9192
08	22 48 08.58790	73 37 02.80601	66.4169
09	22 45 11.27817	73 40 31.15743	84.8819

Consistency of GPS measurement is apparent for point no. 8

Table - 2 Comparison of Coordinates in Everest Datum

The coordinates in first raw are from NRIS database

The coordinates in second raw are from GPS measurement

SR NO	LATITUDE	LONGITUDE	HEIGHT	DIFF. IN LAT.(meters)	DIFF. IN LANG. (maters)	DIFF. IN HEIGHT(meters)
03	22 47 00.292 22 47 00.885	73 33 52.524 73 33 52.481	116.00 110.04	-17.77	01.27	5.96
05	22 47 14.915 22 47 17.722	73 41 04.103 73 41 04.701	155.00 153.42	-24.20	-17.94	1.58
06	22 48 52.992 22 48.53.125	73 33 20.955 73 33 21.492	100.00 110.13	-3.99	-16.08	-10.13
08	22 48 06.375 22 48 07.706	73 37 05.374 73 37 05.685	121.00 121.74	-39.94	-9.29	-0.74
09	22 45 09.741 22 45 10.702	73 40 34.750 73 40 33.678	135.00 139.44	-28.96	32.15	-4.44
13	22 54 35.190 22 54 35.249	73 38 10.065 73 38 11.753	130.00 122.50	-1.77	-52.39	7.5
17	22 50 27.251 22 50 26.968	73 40 27.362 73 40 26.922	140.00 142.88	8.50	13.17	-2.88
50	22 51 43.538 22 51 42.856	73 32 00.678 73 32 02.159	95.00 107.34	20.44	-44.41	-12.34

Differences are within Range of 10-15 meters because of Local Parameters being used for conversion

Table -3 comparison of coordinates in WGS-84 datum

The coordinates in first raw are estimated using Geoid model and seven parameters of transformation for Indian region

The coordinates in second raw are from GPS measurement

SR NO	LATITUDE	LONGITUDE	HEIGHT	DIFF. IN LAT.(meters)	DIFF. IN LANG. (maters)	DIFF. IN HEIGHT(meters)
03	22.78394 22.78389	73.56469 73.56368	70.750 54.868	5.5	108.54	15.88
05	22.78800 22.78792	73.68456 73.68395	109.84 98.448	8.64	65.55	11.39
06	22.81524 22.81498	73.55592 73.55499	54.772 54.543	27.54	100.01	0.23
08	22.80229 22.80245	73.61825 73.61748	75.806 66.417	-17.71	83.37	9.39
09	22.75323 22.75313	73.67640 73.67532	89.803 84.882	10.91	111.29	4.92
13	22.91029 22.91002	73.63622 73.63571	84.915 65.817	29.27	54.65	19.09
17	22.75323 22.75313	73.67641 73.67372	94.881 87.187	30.78	68.69	7.69
50	22.86261 22.86225	73.53363 73.53285	49.798 51.116	39.20	83.59	-1.32