NAME

       x_solve_dc_drift  - Find Best-fitting D.C.-shift and Drift
       Estimates


SYNOPSIS

       x_solve_dc_drift [ -Xxbase ] [ -Llbase ] [ -Iignorefile  ]
       [    -S    ]    [   -Citeration   ]   [   -Mmin_nx   ]   [
       -Rwest/east/south/north ] [  -Uuselegsfile  ]  [  -V  ]  [
       -Bbinfile ] [ -AASCIIfile ]


DESCRIPTION

       x_solve_dc_drift  will  read  the  database  files and, by
       iterating, find the best-fitting regression  line  to  the
       <time,COE>  points for each leg that minimizes the overall
       standard deviation of the data  set  in  a  least  squares
       sense. Finally, correction file(s) are created.

       -X     Indicate an alternate database to read. [Default is
              xx_base.b].

       -L     Indicate an alternate legbase to read. [Default  is
              xx_legs.b].

       -I     Ignore  information for certain legs (that might be
              bad etc).

       -S     Reset the old d.c.-shift  and  drift  estimates  to
              zero before iterating.

       -C     Specify  how  many  iterations  to  do.  Default is
              interactive session.

       -M     Solve for drift only  if  the  leg  has  more  than
              min_nx cross-overs.

       -R     Only  take  COEs  inside  this region into account.
              [Default is world].

       -U     Solve for corrections involving COEs from  legs  in
              the uselegsfile only.

       -B     Create  a binary correction file (which can be read
              by xlist and gmtlist).

       -A     Create an ASCII correction table. At least  one  of
              -A and -B must be specified.

       -V     Operate in verbose mode.


The reason for the -M option is that if the COEs are few, then a

       stable estimate of the slope may
       not be obtained (e.g. when most of the COEs  occur  midway
       between  ports).  However,  in  most cases the bulk of the
       COEs do occur near the ports so that a drift estimate  can
       be  computed.  Conventional wisdom recommends plotting the
       time-series and the computed regression line to see if  it
       makes sense.


SEE ALSO

       gmt(l), x_system(l)


REFERENCES

       Wessel,  P.  XOVER:  A Cross-over Error Detector for Track
       Data, Computers & Geosciences, 15, 333-346.



                           15 Oct 2001        x_solve_dc_drift(l)

Man(1) output converted with man2html