Unconventional applications with single frequency DGPS

RAJNEESH GUPTA, VIVEK BANSAL, BRIG M C DHAMIJA (RETD)

CURRENTLY the trend is that more and more project authorities are mandating use of GPS for control work and private surveyors have to comply with this requirement. Dual frequency DGPS have been beyond the financial reach of these surveyors as such most prefer to invest in single frequency DGPS, which cost almost one third the cost of dual frequency DGPS.

Conventionally DGPS have been used so far for horizontal control. Recently however there have been a number of articles on using GPS for leveling. All the authors have used a dual frequency DGPS. These authors belong to national mapping agencies or academic institutions as such had access to leveling Bench Marks (BMs) of their countries. However, India being a very price sensitive market compounded with the fact, that private surveyors have no access to spirit level height BMs. Even if they work for govt. agencies, obtaining height data involves lot of hassles, security implications, takes unviable time frames and is quite expensive. Finally, when height data is made available, there is no certainty that the BM’s would be found intact on ground. Therefore, the general practice by private surveyors is to make do with alternatives such as commencing leveling from an arbitrary point with arbitrary height value assigned to it. To be competitive in pricing, hardly anybody does a closed loop, nevertheless the work is by and large acceptable by the consultants and project contractors for survey requirements relating to planning, design and checking as built of highways, pipelines, SEZ and so on.

As is well known, DGPS come in two modes i.e. Dual Frequency and Single Frequency. The generally specified accuracies for both are 5mm +1ppm for Horizontal and 10mm +2ppm for vertical in static post processing solutions albeit with a range limit of approx 15 km for single frequency DGPS and upward of 50 km for a dual frequency DGPS. The versatility of both considerably differs for various applications.

Single frequency DGPS can also be used in Static PP, Stop & GO Kinematic Post Processed (PP) and continuous Kinematic PP. Besides single frequency DGPS can also be used for GIS data acquisition with PP accuracy of submeter for Base lines up to 50 km.

Background

Many of our colleagues in the Institution of Surveyors (India) and private surveyors had been querying us to suggest methodology for utilizing single frequency DGPS for extended range to cover their projects involving large distances. This article is an endeavor to present some experiments with single frequency DGPS both for distance as well
elevation accuracies over large distances.

Pan India has been an accredited in house R&D agency by the Dept. of Scientific & Industrial Research, Ministry of Science & Technology, Govt. of India. Pan India in its ambit of R&D activity has undertaken a detailed study to examine some of the unconventional solutions for surveys using a single frequency DGPS in project spanning distances over 100-150 km.

However, before undertaking investigations we had a wide ranging discussion with various experts, it has been our experience that there are mainly 3 classes of users, the first is an organization such as Survey of India (SOI), who are very well versed with equipment as well as technology and surveying practices.

The second category of surveyors are those such as M/s Punj Lloyd / M/s Jaypee etc. who’s main job is construction related and surveying is a essential component mainly for acquiring Base data for design, subsequently for ground layout and finally for checking as built. The projects are related to Highways, Pipelines and layouts for townships or SEZ etc.

The third category of users are private surveying companies who generally work as sub-contractors for survey works for infrastructure companies and consultants undertaking preparation of DPRS designs etc.

It is for benefit of last 2 categories of surveyors that we had undertaken different

projects to evaluate the utility of single frequency DGPS beyond the range of12-15 km. These survey companies need establishment of control over a distance of 100 to 200 km for highways, L Section and Cross section, layout surveys for pipelines extending over distances in excess of 100 km, DEM/DTM for town planning & SEZ’s and soon.

The project

Four different types of projects were selected for the unconventional procedures undertaken for evaluation of utility of single frequency DGPS as follows:
1) Control work for highway alignment covering a distance of over 118 km in UP.
2) L section & Cross section survey for about 2 km.
3 ) Data acquisition for DEM for 20 Acres Part of SEZ.
4 ) As built survey of 130 km part of Oil pipe line.

The procedure

As already mentioned the Single Frequency DGPS has optimum range of to 12-15 km to achieve specified accuracy in post processing. In order to increase this range the obvious solution is to use Base and Rover in leap-frog mode i.e. first set up DGPS on Base and Rover on stations about 8-10 km apart. Next whilekeeping the Rover intact, remove Base GPS and put it on to the next station as Rover 8-10 km further away, in this way the DGPS unit acting as Rover earlier will now act as Base. The control points were thus extended in similar way by making leap frog point between 6 to 11 km apart for the complete route of about 120 km. During post processing, it was made mandatory that data for 1st Base & 1st Rover position was processed. In the second Base-Rover processing the Base (which was Rover in first observation) was assigned the same coordinates derived from earlier processing & so on. For automatic processing in similar way for all position the utility existing in the software though never used earlier, was utilized. With this utility, the processing is ordered in the desired sequencing of Base-Rover-Base with automatic assigning of processed co-ordinates which were obtained for the position as earlier Rover to that of Base in subsequent processing to follow the procedure as mentioned above. In a route spanning 120 km a total of 12 stations were established in this manner (see diag 1).

Results

Extension of Control Points

To check the over all accuracy of results, a single quadrilateral was observed over entire distance using Dual Frequency DGPS and processed by trilateration. The initial azimuth was taken as the one between Base & 1st Rover. The closing azimuth was checked with the last Base & Rover station. The averaged value of coordinates of 1st Base observed over a period of 2 hours was taken as starting co-ordinate in UTM zone 43 and WGS-84 Datum.

The results on comparison showed that the distance vector between end points, which were 107 km direct devised by single frequency GPS in Leap frog mode are in agreement with in 4 ppm. The azimuth is in agreement in less than one second. The ellipsoidal elevation by both single frequency DGPS at 1st Base and last Rover (approx 107 km direct) & (118 kmby leap-frog) were in agreement to with in single tertiary Class II specification as compared with Dual Frequency.

Since absolute MSL height of starting & end point was not available, EGM 96 was applied to obtain Orthometric Heights. Comparison between Δh and ΔH was done to study the variation in ΔN with respect to EGM-96 which showed that the value of N in this project area was varying between -0.269 to + 0.257 over a distance of 118 km in generally fl at terrain (table 1).

Subsequently a BM found near point 401 which is a canal BM and second near point 413 which is a railway BM. These are in different series and by different departments. However when the heights were observed at these points by DGPS and substituted for point 401 instead of value obtained in EGM 96 and with this height fixed, all other points ware datumed on to it.. The value of height of railway BM at point 413 as compared with the height inscribed on the railway BM was found to be in agreement in better than 47 cms in the leap-frog series (table-2).

RAJNEESH GUPTA, VIVEK BANSAL, BRIG M C DHAMIJA (RETD)

Even though apparently it falls short of double tertiary class II standard, the fact remains that the average correction per linear km is less than 5 mm and per leg of average distance 8 km (i.e. 107÷13) the error is with in 6 cms for each leg of 8 km, which after adjustment where accurate BM heights are known is acceptable for most surveys relating to highways and pipeline & other infrastructure surveys as well as are adequate for topo surveys. Where VI for contouring is 1 meter or more.

DEM/DTM

Under normal conditions, Single Frequency DGPS stipulates that 15 epoch acquisition of data is adequate to achieve the specified accuracy in X, Y & Z. For any DEM / DTM all threevalues of X, Y & Z are necessary. In our trials, an area of about 20 Acres was taken up for preparation of DEM / DTM. Single frequency DGPS was used in Stop & Go mode for acquisition of xyz for points spaced approx every 5 meters. The Base was established in centre of area and Rovers were initialized using initializing bar. The recording interval was kept at one second so that time for data acquisition at each point is not more than 20 seconds. Besides one Base station, 3 Rovers were utilized for data acquisition in grid pattern. In this configuration, the XYZ for approx 3000 points acquired in one day, which was considered to be very economical out turn as compared with other conventional instruments. Alongwith the grid pattern data acquisition, the surveyors picked up positions of breakpoints as well as features of interest.
Repeat observation was taken on well spread 50 points for comparison. The average ellipsoidal height discrepancy on the repeat points was with in 1cm and the maximum discrepancy in repeat point at extreme ends was with in 2 cms. Standard software with 8 degree polynomial was used for drawing DEM. A check line at random was carried out which showed that all elevations were in agreement with in 1.5 cm.

L Section & Cross Section

In the third trial one L section of 500 meters with cross section at every 50 meters was observed by single frequency DGPS in stop & go Kinematic mode and post processed part of this was checked using the
auto level. The discrepancy in relative ellipsoidal heights for this small area was maximum of 3mm at the extremity.

As Built Survey

The fourth project involving survey of ‘As Built’ of oil pipeline was carried out by surveyors M/s Punj Lloyd for their project Dahej-Uran gas Pipeline (DUPL) for GAIL. The pipeline had been laid over a distance of 132 km with turns, curves, and river crossings. The work involved in co-ordinating turns as well as level grading that the as built pipelines adhered to. This was successfully carried out using single frequency DGPS in leap-frog mode by surveyors of M/s Punj Lloyd with our active involvement and under the continued monitoring by their consultant and resulted in considerable time saving, as compared to the conventional systems.

Discussion

In the absence of MSL bench marks, all elevations over longer distance were converted to orthometric heights using EGM-96. However where ever SOI leveling bench marks are made available a suitable Geoid model with interpolation based on suitable polynomial curve fitting should meet the specifications of project indenters. Presently hardlyany contactor has access to SOI bench marks and they usually commence their work from an arbitrarily established point. However by & large accuracy of single tertiary class II suffices for their projects. Accordingly they are able to meet their client’s requirements with relative difference of ellipsoidal heights with EGM96 orthometric heights datumed to mean value of elevation of beginning point which is derived from SOI TOPO maps, so that the elevation data is not too much in variation with absolute M.S.L. elevation in the project area. Hopefully, this situation would improve with technology awareness and opening up measures that SOI has initiated for data availability as well as availability of INDGEOID on which G&RB, Survey of India is currently working. We hope this Geoid model will be available to private surveyors as well.

The 118 km road line was completed in 3 working days. The DEM observations and processing was completed in 2 days and the L section X section was completed in one working day. The asbuilt pipe line of 130 km (approx) was completed in about 25 days under constant scrutiny of the consultant’s engineers.

Concluding remarks

We may also add a word of caution that as far as possible in road/canal/pipeline projects in generally fl at areas the GPS derived heights should as a rule be done section by section of limited lengths.
In extremely fl at areas where the slope measurement is very critical to the project a few small sections of spirit leveling as checkup may be done. One must make appreciation of the tolerance threshold and if this threshold is single tertiary class II, the relative height information can be confidently derived by single frequency DGPS, in leap frog procedure.

Single Frequency DGPS are very versatile, handy and give precise results. The cost of ownership is much lower than those of Dual Frequency as such ROI is realizable in just one or two survey projects. Dual frequency DGPS of course have their own applications.

Acknowledgements

The initial inspiration & guidance for undertaking this study was provided by Brig (Dr.) B. Nagarajan, Director, Geodetic & Research Branch, Survey of India.

The second impetus has been various articles & regular column in various issues of Co-ordinates by Dr. Muneendra Kumar, Chief Geodesist (Retd.) USNGIA. Dr. Kumar has been a strong proponent of utilizing ellipsoidal heights for engineering and Topo survey projects.

The acknowledgement can not be complete without mentioning Prof. P. Misra whose pearls of wisdom exhorts that merely having high level of knowledge of technology is of no help to any body. The challenge before professionals is to be aware of the problem solving abilities of technology and then utilize it for the same.

References

1 ) Redefinition of Indian Geodetic datums Horizontal and Vertical, Brig (Dr.) B. Nagarajan, R.K. Sawhney, Coordinates Vol. III issue 4 April 2007
2 ) Determining of Local gravimetric Geoid, S.K. Singh, Brig (Dr) B. Nagarajan, P.K. Garg, Co-ordinates Vol. III issue 2, Feb 2007.
3 ) Geoid Vol. III issue, Jan-March 2007.
4 ) GPS Based control points for mapping, Jayanta Kumar Ghosh, Ojaswa Sarmg, Amit Goyal,l 5 ) Co-ordinates Vol. II issue 10 Oct 2006.A Cost Effective GPS leveling method versus conventional method
for typical surveying application. Mr. Metin Soycan, GIS Development Asia Pacific, Vol 10 issue 8 Aug 2006
6 ) Ellipsoidal heights and engineering applications, Muneendra Kumar Ph.D, Co-ordinates Vol. III issue 1, Jan 2006.
7 )Global usage of Ellipsoidal Heights Dr. Muneendra Kumar, Coordinates Vol. issue II July 2005
8 ) Orthometric Heights From GPS-Leveling Observations, Brig. (Dr.) B. Nagarajan, S.K. Singh, GIS Development Vol. 9 issue 4 April 2005.