GISdevelopment.net --> Policy --> Geographic Information Infrastructure


Design of Large Area Spatial Framework for National (Natural) Resources Information System

R K Goel
Head, Geomatics Technology Division Space Applications Centre Ahmedabad – 380053


For the sustainable development and monitoring of natural resources , Indian Space Research Organization has started a major program National (Natural) Resources Information System (NRIS). NRIS is visualized as network of Computerized, Integrated (Spatial & Non-Spatial) databases around GIS as core for handling Natural Resources, Infrastructure and Socio-economic data at District, State and Center level. Design and implementation of suitable mechanisms for sharing and transfer of data across district, state and center nodes is of utmost importance for this program. India being a large Geographic entity, design and implementation of large area spatial frame work is a pre-requisite for realizing computerized mechanism which can facilitate bottom-up and horizontal transfers and aggregation of data amongst the various District, State and central nodes..

In order to test different projection/coordinate system, controlled spatial features are created in topomaps of extreme regions of India. These features are then projected at corresponding district, state and center level. Results obtained for each projection are compared in terms of distortions in area, perimeter and length of the features with respect to the original values at district level. Projection is accepted if distortions are less than the tolerance limits specified in NRIS Standards Document [1].

This technical paper presents the results of a systems study conducted and suitability of projection and co-ordinate system to be adopted for creating large area spatial framework for NRIS node network

Selection of Projections for the study
The projections are created by projecting three dimensional surface of the earth on the surfaces like plane, cone or cylinder and these surfaces are developed into a two dimensional plane. Because of reduction in dimensionally and approximation of the earth into planar, conical or a cylindrical surface geometric properties of features like shape, area, scale and direction(angle) change and distortions in all or some of the geometric properties are introduced depending upon the type of projection.

Since the NRIS database has to be created at India level covering large area , it becomes necessary that the projection system selected introduces minimum or tolerable distortions in geometrical properties while aggregating spatial data from district level to state level and from state level to centre level. For this purpose , the projection systems having following criteria are considered for the study.

Criteria for selecting appropriate projection
  • Suitability for creating seamless data base at all India level.
  • Regional characteristics viz. east-west and north-south.
  • Geometric properties viz. shape ,area ,scale and direction.
  • Suitability for thematic and topographic analysis and presentation.
  • Suitability for mapping at state level (medium area) and at center level (large area).
  • Suitability for data aggregation at state and center level.
Based on above criteria three projection systems viz. Polyconic , Transverse Mercator and Lambert conformal conic with two standard parallels are selected for the study. The UTM projection is not selected because of its limitation due to its Zonal characteristic to represent only 6 degree longitude by 8 degrees latitudes area. The major properties of these projections as described in Map Projection manual [2] are as follows:

Polyconic Projection : Created by lining up infinite number of cones along central meridian where meridians are complex curve and parallels are not concentric circles. But as a compromise meridians/parallels are represented as straight line segments. It introduces minimum distortions in shape , area, scale and direction near central meridian and distortions increase with respect to. distance from central meridian. Local angles are accurate along the central meridian. The scale along each parallel and along the central meridian is accurate. It increases along the meridians as the distance from the central meridian increases One reason for selecting this projection is that SOI topographic maps are prepared using Polyconic projection.

Transverse Mercator Projection : This is a cylindrical ,conformal projection. Cylinder is kept longitudinal along meridian instead of along the equator. The central meridian is centered on the region to be projected therefore all distortion are minimum along north-south direction because meridians run north-south. Small shapes are maintained but larger shapes are distorted while going away from the central meridian. Since UTM can not be considered for representing entire India his projection is selected for the study.

Lambert Conformal Conic Projection :
This conic projection uses two standard parallels to reduce some of the distortion produced when only one standard parallels is used. The distortion related to shape and scale are minimum in region between two standard parallels. The spacing between the lines of latitude increases beyond the standard parallel .Suitable for middle latitude region Local angles are accurate throughout because of conformity. scale is correct along the standard parallels and is reduced between parallels and increases beyond the parallels.

Methodology
Four regions pertaining to topomaps 41E , 53E , 58F and 73M along east, west , north and south extremes of INDIA were selected. Fictitious line and polygon features were created in geographic projection system using ARCEDIT functionality of ARC/INFO. For each map twelve projection files were created and using these projection files the four basic coverages were projected considering district ,state and center meridian for the selected Polyconic ,Lambert conformal conic and Transverse Mercator projection system. For analytic comparison, Polygon Attribute Table (PAT) files containing area and perimeter of polygon features and Arc Attribute Table (AAT) files containing length of line features were used. Details of distortions in area and length of features at state node and at centre node are given in table.

Results and Conclusion
An attempt is made to estimate errors introduced in geometric properties of the features like area, perimeter and length due to shift in center of projection from topomap to district, district to state and district to center level for Polyconic , Lambert conformal conic and Transverse Mercator projections. Tolerable limit of errors specified in NRIS Standards Document [1] is 0.3 % of features area ,perimeter or length. Projection parameters and Results of study are given in Table 1and Tables 2A, 2B , 2C.

It is observed that while shifting projection center from topomap level to district level , Polyconic and Transverse Mercator projections do not show any error where as Lambert projection shows very negligible error which is less then 0.003 % in area , perimeter and length of the features. Hence , all of above projections are suitable at district level.

For district and state level, errors are within tolerable limits for all of above projections but Lambert Conformal Conic projections should be preferred due to its conformal characteristic. For the centre level no projection is found suitable to create seamless database .

Future Work
This study has evaluated the three projections viz. Polyconic, Transverse Mercator and Lambert conformal conic and found that none of these projections are suitable for creating seamless database at center level. Therefore Geographic projection or grid based projection /coordinate system should be adopted for creating seamless database at centre level.

Acknowledgements The authors are thankful to Shri A.R. Dasgupta, Project Director, NRIS-SIP and Dy. Director, SIIPA, for the constant guidance and support provided by him for carrying out this study.

Referencers

  • Node Design and Standards for NRIS ,July 1998. NO. SAC/SIIPA/NRIS-SIP/SD-02/98
  • Map Projections manual by ESRI.
Table 1: Projection Centres for Different Maps at District, State and Centre Level
MAP No. DISTRICT LEVEL Longitude Latitude STATE LEVEL Longitude Latitude CENTER LEVEL Longitude Latitude PROJECTION
41E 69 30 23 30 71 30 22 45 80 00 21 00 Polyconic/ Mercator
" 69 30 23 10 23 50 71 30 20 45 23 45 80 00 13 00 29 00 Lambert
53E 77 30 31 45 77 00 32 00 80 00 21 00 Polyconic/ Mercator
77 30 31 25 32 05 77 00 31 10 31 50 80 00 13 00 29 00 Lambert
58F 77 15 10 45 78 15 10 45 80 00 21 00 Polyconic/ Mercator
77 15 10 10 10 50 78 15 09 50 11 10 80 00 13 00 29 00 Lambert
73M 87 30 23 30 87 30 23 30 80 00 21 00 Polyconic/ Mercator
87 30 23 10 23 50 87 30 22 20 23 40 80 00 13 00 29 00 Lambert


Table 2A : Average Distortions at State and Centre Level with Respect to District Level for Polyconic Projection
Map Distortions at State Level Distortions at Central Level
No. Area Perimeter Length Area Perimeter Length
  sq. km. % m % m. % sq. km. % m % m %
41E 0.27 0.052 25.5 0.028 3 0.008 7.01 1.42 799 0.93 131 0.348
53E 0.01 0.004 1.5 0.002 1 0.003 0.33 0.069 35.3 0.04 5 0.014
58F 0.01 0.006 4.75 0.005 0.5 0.002 0.51 0.091 45.3 0.048 6 0.015
73M 0.0 0.0 0.0 0.0 0.0 0.0 3.47 0.724 362 0.402 82 0.218



Table 2B: Average Distortions at State and Centre Level with Respect to District Level for Transverse Mercator Projection

Map Distortions at State Level Distortions at Central Level
No. Area Perimeter Length Area Perimeter Length
sq. km. % m % m. % sq. km. % m % m %
41E 0.38 0.103 47.5 0.051 10 0.031 2.07 0.410 186.7 0.204 52 0.165
53E 0.04 0.006 4.5 0.003 2 0.006 0.667 0.140 53 0.060 14 0.047
58F 0.115 0.019 9.25 0.001 1.5 0.004 1.01 0.185 86.5 0.090 20 0.061
73M 0.0 0.0 0.0 0.0 0.0 0.0 6.99 1.462 1157 1.331 250 0.788


Table 2 C: Average Distortions at State and Centre Level with Respect to District Level for Lambert Conic Projection
Map Distortions at State Level Distortions at Central Level
No. Area Perimeter Length Area Perimeter Length
sq. km. % m % m. % sq. km. % m % m %
41E -0.18 -0.04 -16.8 -0.02 -3.5 -0.01 -23.8 -4.88 -2235 -2.47 -775 -2.46
53E -0.07 -0.02 -7.75 -0.01 -3.0 -0.01 -8.3 -1.83 -802 -1.18 -112 -0.47
58F -0.12 -0.02 -11.0 -0.01 -3.5 -0.01 -9.2 -1.74 -822 -0.88 -312 -0.93
73M -0.18 -0.02 -6.75 -0.01 -1.5 -0.01 -23.8 -4.88 -2235 -2.47 -775 -2.46


The highlighted figures of % error indicate that these errors are greater than acceptable limit of 0.3%
© GISdevelopment.net. All rights reserved.