ESTPM

ESTPM was been developed to estimate corrections to geodetic positions by interpolating between the position corrections which are known for a number of control stations. ESTPM estimates position corrections at any location within the periphery of control stations with known position corrections. The estimation is done first at the datum level by means of geometric coordinate transformations, then at the regional level with a least-squares complex polynomial for regional trends, and finally at the local level using a covariance function for residual interpolation.

All the matrix and vector dimensions are computed within the program at execution time depending on the amount of data read in and the chosen degree for the complex polynomial. When insufficient central memory is available, a diagnostic message is printed in the output listing giving the amount of additional core required and the execution is aborted.

DOCUMENTATION

Program ESTPM was originally written by J.A.R. BLAIS, July 1979 Geodetic Survey Division, Surveys and Mapping Branch Department of Energy Mines and Resources, Canada

Documentation added and modifications by G.A.Wilkinson, July 1986 Surveys and Resource Mapping Branch British Columbia Ministry of Environment

ESTIMATION OF SECONDARY TERRESTRIAL POSITIONS FOR MAPPING

ESTPM was been developed to estimate corrections to geodetic positions by interpolating between the position corrections which are known for a number of control stations. ESTPM estimates position corrections at any location within the periphery of control stations with known position corrections. The estimation is done first at the datum level by means of geometric coordinate transformations, then at the regional level with a least-squares complex polynomial for regional trends, and finally at the local level using a covariance function for residual interpolation.

The datum transformation is optional and a number of datum transformations are available.

At the regional level, a complex polynomial model is used to represent the regional trends in the given correction vectors at the control stations. For a given degree, the complex coefficients of the polynomial are estimated by the method of least-squares. There also is provision for situations where the input degree is too high for the input observational information so that only the lower-order coefficients numerically feasible are estimated and the higher-order coefficients are set to zero.

At the local level the remaining residual corrections at the control stations can be practically eliminated through interpolation with an appropriately calibrated covariance function. The covariance function used in ESTPM is the inverse exponential of the distance, which corresponds to a Gaussian stochastic process. The calibration of the covariance function is done using the input normalized unit distance (RUI) in terms of the spatial dispersion of the control stations. The default value for the normalized unit distance is one tenth of the radial dispersion of the control stations about their centre of gravity.

The analysis of ESTPM results may be simplified by use of the optional graphical outputs. The given correction vectors at the control stations and the estimated correction vectors at grid and other points provide a useful graphical representation of the results for assessment and analysis purposes. These outputs can be displayed using an interactive graphical computer terminal or a CALCOMP plotter.

The estimated coefficients and related information can also be saved or catalogued for future uses. These include the transformation of positions at a later date and/or further graphical analysis especially in overlapping areas.

The optional graphical output is generated using the standard CDC FORTRAN calls to CALCOMP routines for PLOT, AXIS, SYMBOL, and so on. The interactive use ot the program may necessitate minor changes to those CALL statements to conform with the interactive system software.

The system SORT/MERGE is used to find the matching stations on the input files IT2 and IT3. Out of those matching stations, only those within the input grid limits are kept for the modelling. Note that the grid limits are always required for this purpose even when no grid is wanted. The resulting transformations are then applied only to those control stations and the other stations on IT4 within the area defined by the peripheral control stations and the parameter RUE.

REFERENCES

Blais, J.A.R. (1979): Least-Squares Estimation of Geodetic Horizontal Control Densification. Presented at the Canadian Geophysical Union Meeting at UNB (June 1979).

Boal, J.D. and Junkins, D.R. (1978): Adjustment of Inertial Survey System Data with Satellite Doppler and Conventional Control. Collected Papers (1978), Geodetic Survey of Canada.

CONTROL_FILE

The control information required by ESTPM is read in from two cards:

  Card 1: Project Title
  Columns 1-80  TITLE  Reproduced without any modification in the output
                       listings and on the plot for identification purposes.
                       (a80)

  Card 2: Control Parameters
  Column
   1- 2  NS   Degree of complex polynomial requested - No Default
              99 --> coefficient data is to be read on Unit 11
      3  K1   Option variable for datum transformations
              to be applied to coordinates on IT3 and IT4  (i1)
              0 - no datum transformation - Default
              1 - transformation NAD27->NAD83
              2 - transformation NAD83->NAD27
              3 - transformation MAY76->NAD83
              4 - transformation NAD83->MAY76
              5 - transformation NAD27->MAY76
              6 - transformation MAY76->NAD27
              7 - transformation ATS77->NAD83
              8 - transformation NAD83->ATS77
              9 - transformation ATS77->MAY76
      4  K2   Plot option (i1)
              0 - no plot - Default
              1 - plot with point numbers
              2 - plot without point numbers
      5  K3   Grid option (i1)
              0 - no grid - Default
              1 - usual grid
              2 - grid symbols without vectors
   6-10  GW1   Eastern longitude limit in degrees (f5.1) - No Default
  11-15  GW2   Western longitude limit in degrees (f5.1) - No Default
  16-20  GN1   Southern latitude limit in degrees (f5.1) - No Default
  21-25  GN2   Northern latitude limit in degrees (f5.1) - No Default
  26-30  SGD   Plot Scale in degrees of latitude per inch (f5.3)
               Default = 1 degree per inch
  31-35  SCV   Plot Scale of vectors in seconds per inch (f5.3)
               Default = 1 second per inch
  36-40  GRD   Spacing of grid in degrees (f5.3)
               Default = 1 degree spacing
  41-45  RCV   Maximum vector length in inches (f5.3)
               Default = 3.0 inches
  46-50  SBT   Size of title letters in inches (f5.3)
               Default = 0.21 inches
  51-55  SBS   Size of symbols in inches (f5.3)
               Default = 0.14 inches
  56-60  SBI   Size of point numbers in inches (f5.3)
               Default = 0.07 inches
  61-65  RUI   Normalized unit distance in terms of radial dispersion
               of control stations with respect to their centre of
               gravity.  Used for interpolation by the covariance
               function (f5.3).
               Negative value --> no local interpolation.
               Default = 0.1 (of radial dispersion)
  66-70  RUE   Normalized unit distance in terms of radial dispersion
               of control stations with respect to their centre of
               gravity for extension beyond the peripheral control
               stations (f5.3).
               Default = 0.1 (of radial dispersion)
     71  K4    Option code for output of non-coordinate records
               Blank --> Output non-coordinate records
               NonBlank --> Do not output non-coordinate records
     80  KF    Control code for data formats
               Blank --> GALS format for data cards.
               NonBlank --> Read Format on following card (a80)
               Z --> No checking for position codes 4, 5, 6, 04, 05, 06,
                     and 96. All records assumed to be coordinate records

  Card 3: Optional data format read only if Column 80 on Card 2 is non-blank.
  Columns 1-80  FMT   Data format to be used to read coordinate records (a80)
 

DATA_FILES

     INPUT       !Input
     OLDADJ      !First set of coordinates of control stations
                 !Not required when NS = 99
     NEWADJ      !Second set of coordinates of control stations
                 !Not required when NS = 99
     OLDPOS      !Coordinates of stations to be transformed
     NEWPOS      !Transformed Coordinates of stations from OLDPOS
     IPT         !Plot file output
     ESTPM.LIS   !Output
     COEFOT.DAT  !ASCII  file of polynomial coefficients and related data.
                 !Required only when (degree .eq. 99)
                 !Created by ESTPM when (degree .ne. 99)
     ITQ         !Temporary data file
                 !Used to store coordinates in IT2 & IT3 for
                 !Input to SORTFILE, called within DATAST
     ITR         !Temporary sort file. Used only in DATAST.
                 !Output from SORTFILE
     ITS         !Temporary file for positions and position
                 !differences.  Written in DATAST and
                 !read in LSQSLN.

INTERACTIVE_PROCEDURE

The user is queried for the type of job:

BATCH_PROCEDURE

For a batch job the user must include parameters as follows:

The batch submission command would look like:

   $ SUBMIT/parameter=(p1,p2,p3,p4,p5,p6,p7) ghost$cmnd:estpm

All files must have the extension .DAT.

The extension must not be entered in the parameter list.

All parameters must be entered and must be in UPPER CASE letters.