1SURFACE(1) Generic Mapping Tools SURFACE(1)
2
6 surface - adjustable tension continuous curvature surface gridding
7 algorithm
8
10 surface [ xyzfile ] -Goutputfile.grd
11 -Ixinc[unit][=|+][/yinc[unit][=|+]] -Rwest/east/south/north[r] [
12 -Aaspect_ratio ] [ -Cconvergence_limit ] [ -H[i][nrec] ] [ -Lllower ] [
13 -Luupper ] [ -Nmax_iterations ] [ -Q ] [ -Ssearch_radius[m] ] [ -Tten‐
14 sion_factor[i|b] ] [ -V[l] ] [ -Zover-relaxation_factor ] [ -:[i|o] ] [
15 -bi[s|S|d|D[ncol]|c[var1/...]] ] [ -fcolinfo ]
16
18 surface reads randomly-spaced (x,y,z) triples from standard input [or
19 xyzfile] and produces a binary grid file of gridded values z(x,y) by
20 solving:
21 (1 - T) * L (L (z)) + T * L (z) = 0
23 where T is a tension factor between 0 and 1, and L indicates the Lapla‐
25 cian operator. T = 0 gives the "minimum curvature" solution which is
26 equivalent to SuperMISP and the ISM packages. Minimum curvature can
27 cause undesired oscillations and false local maxima or minima (See
28 Smith and Wessel, 1990), and you may wish to use T > 0 to suppress
29 these effects. Experience suggests T ~ 0.25 usually looks good for
30 potential field data and T should be larger (T ~ 0.35) for steep topog‐
31 raphy data. T = 1 gives a harmonic surface (no maxima or minima are
32 possible except at control data points). It is recommended that the
33 user pre-process the data with blockmean, blockmedian, or blockmode to
34 avoid spatial aliasing and eliminate redundant data. You may impose
35 lower and/or upper bounds on the solution. These may be entered in the
36 form of a fixed value, a grid with values, or simply be the mini‐
37 mum/maximum input data values.
38 xyzfile
40 3 column ASCII file [or binary, see -b] holding (x,y,z) data
41 values. If no file is specified, surface will read from stan‐
42 dard input.
43 -G Output file name. Output is a binary 2-D .grd file. Note that
45 the smallest grid dimension must be at least 4.
46 -I x_inc [and optionally y_inc] is the grid spacing. Optionally,
48 append a suffix modifier. Geographical (degrees) coordinates:
49 Append m to indicate arc minutes or c to indicate arc seconds.
50 If one of the units e, k, i, or n is appended instead, the
51 increment is assumed to be given in meter, km, miles, or nauti‐
52 cal miles, respectively, and will be converted to the equivalent
53 degrees longitude at the middle latitude of the region (the con‐
54 version depends on ELLIPSOID). If /y_inc is given but set to 0
55 it will be reset equal to x_inc; otherwise it will be converted
56 to degrees latitude. All coordinates: If = is appended then the
57 corresponding max x (east) or y (north) may be slightly adjusted
58 to fit exactly the given increment [by default the increment may
59 be adjusted slightly to fit the given domain]. Finally, instead
60 of giving an increment you may specify the number of nodes
61 desired by appending + to the supplied integer argument; the
62 increment is then recalculated from the number of nodes and the
63 domain. The resulting increment value depends on whether you
64 have selected a gridline-registered or pixel-registered grid;
65 see Appendix B for details. Note: if -Rgrdfile is used then
66 grid spacing has already been initialized; use -I to override
67 the values.
68 -R xmin, xmax, ymin, and ymax specify the Region of interest. For
70 geographic regions, these limits correspond to west, east,
71 south, and north and you may specify them in decimal degrees or
72 in [+-]dd:mm[:ss.xxx][W|E|S|N] format. Append r if lower left
73 and upper right map coordinates are given instead of w/e/s/n.
74 The two shorthands -Rg and -Rd stand for global domain (0/360
75 and -180/+180 in longitude respectively, with -90/+90 in lati‐
76 tude). Alternatively, specify the name of an existing grid file
77 and the -R settings (and grid spacing, if applicable) are copied
78 from the grid. For calendar time coordinates you may either
79 give (a) relative time (relative to the selected TIME_EPOCH and
80 in the selected TIME_UNIT; append t to -JX|x), or (b) absolute
81 time of the form [date]T[clock] (append T to -JX|x). At least
82 one of date and clock must be present; the T is always required.
83 The date string must be of the form [-]yyyy[-mm[-dd]] (Gregorian
84 calendar) or yyyy[-Www[-d]] (ISO week calendar), while the clock
85 string must be of the form hh:mm:ss[.xxx]. The use of delim‐
86 iters and their type and positions must be exactly as indicated
87 (however, input, output and plot formats are customizable; see
88 gmtdefaults).
89
91 -A Aspect ratio. If desired, grid anisotropy can be added to the
92 equations. Enter aspect_ratio, where dy = dx / aspect_ratio
93 relates the grid dimensions. [Default = 1 assumes isotropic
94 grid.]
95 -C Convergence limit. Iteration is assumed to have converged when
97 the maximum absolute change in any grid value is less than con‐
98 vergence_limit. (Units same as data z units). [Default is
99 scaled to 0.1 percent of typical gradient in input data.]
100 -H Input file(s) has header record(s). If used, the default number
102 of header records is N_HEADER_RECS. Use -Hi if only input data
103 should have header records [Default will write out header
104 records if the input data have them]. Blank lines and lines
105 starting with # are always skipped. Not used with binary data.
106 -L Impose limits on the output solution. llower sets the lower
108 bound. lower can be the name of a grid file with lower bound
109 values, a fixed value, d to set to minimum input value, or u for
110 unconstrained [Default]. uupper sets the upper bound and can be
111 the name of a grid file with upper bound values, a fixed value,
112 d to set to maximum input value, or u for unconstrained
113 [Default].
114 -N Number of iterations. Iteration will cease when conver‐
116 gence_limit is reached or when number of iterations reaches
117 max_iterations. [Default is 250.]
118 -Q Suggest grid dimensions which have a highly composite greatest
120 common factor. This allows surface to use several intermediate
121 steps in the solution, yielding faster run times and better
122 results. The sizes suggested by -Q can be achieved by altering
123 -R and/or -I. You can recover the -R and -I you want later by
124 using grdsample or grdcut on the output of surface.
125 -S Search radius. Enter search_radius in same units as x,y data;
127 append m to indicate minutes. This is used to initialize the
128 grid before the first iteration; it is not worth the time unless
129 the grid lattice is prime and cannot have regional stages.
130 [Default = 0.0 and no search is made.]
131 -T Tension factor[s]. These must be between 0 and 1. Tension may
133 be used in the interior solution (above equation, where it sup‐
134 presses spurious oscillations) and in the boundary conditions
135 (where it tends to flatten the solution approaching the edges).
136 Using zero for both values results in a minimum curvature sur‐
137 face with free edges, i.e. a natural bicubic spline. Use -Tten‐
138 sion_factori to set interior tension, and -Ttension_factorb to
139 set boundary tension. If you do not append i or b, both will be
140 set to the same value. [Default = 0 for both gives minimum cur‐
141 vature solution.]
142 -V Selects verbose mode, which will send progress reports to stderr
144 [Default runs "silently"]. -Vl will report the convergence
145 after each iteration; -V will report only after each regional
146 grid is converged.
147 -Z Over-relaxation factor. This parameter is used to accelerate
149 the convergence; it is a number between 1 and 2. A value of 1
150 iterates the equations exactly, and will always assure stable
151 convergence. Larger values overestimate the incremental changes
152 during convergence, and will reach a solution more rapidly but
153 may become unstable. If you use a large value for this factor,
154 it is a good idea to monitor each iteration with the -Vl option.
155 [Default = 1.4 converges quickly and is almost always stable.]
156 -: Toggles between (longitude,latitude) and (latitude,longitude)
158 input and/or output. [Default is (longitude,latitude)]. Append
159 i to select input only or o to select output only. [Default
160 affects both].
161 -bi Selects binary input. Append s for single precision [Default is
163 d (double)]. Uppercase S or D will force byte-swapping.
164 Optionally, append ncol, the number of columns in your binary
165 input file if it exceeds the columns needed by the program. Or
166 append c if the input file is netCDF. Optionally, append
167 var1/var2/... to specify the variables to be read. [Default is
168 3 input columns].
169 -f Special formatting of input and/or output columns (time or geo‐
171 graphical data). Specify i or o to make this apply only to
172 input or output [Default applies to both]. Give one or more
173 columns (or column ranges) separated by commas. Append T (abso‐
174 lute calendar time), t (relative time in chosen TIME_UNIT since
175 TIME_EPOCH), x (longitude), y (latitude), or f (floating point)
176 to each column or column range item. Shorthand -f[i|o]g means
177 -f[i|o]0x,1y (geographic coordinates).
178
180 Regardless of the precision of the input data, GMT programs that create
181 grid files will internally hold the grids in 4-byte floating point
182 arrays. This is done to conserve memory and furthermore most if not
183 all real data can be stored using 4-byte floating point values. Data
184 with higher precision (i.e., double precision values) will lose that
185 precision once GMT operates on the grid or writes out new grids. To
186 limit loss of precision when processing data you should always consider
187 normalizing the data prior to processing.
188
190 To grid 5 by 5 minute gravity block means from the ASCII data in
191 hawaii_5x5.xyg, using a tension_factor = 0.25, a convergence_limit =
192 0.1 milligal, writing the result to a file called hawaii_grd.grd, and
193 monitoring each iteration, try:
194 surface hawaii_5x5.xyg -R198/208/18/25 -I5m -Ghawaii_grd.grd -T0.25
196 -C0.1 -Vl
197
199 surface will complain when more than one data point is found for any
200 node and suggest that you run blockmean, blockmedian, or blockmode
201 first. If you did run blockm* and still get this message it usually
202 means that your grid spacing is so small that you need more decimals in
203 the output format used by blockm*. You may specify more decimal places
204 by editing the parameter D_FORMAT in your .gmtdefaults4 file prior to
205 running blockm*, or choose binary input and/or output using single or
206 double precision storage.
207 Note that only gridline registration is possible with surface. If you
208 need a pixel-registered grid you can resample a gridline registered
209 grid using grdsample -T.
210
212 blockmean(1), blockmedian(1), blockmode(1), GMT(1), nearneighbor(1),
213 triangulate(1)
214
216 Smith, W. H. F, and P. Wessel, 1990, Gridding with continuous curvature
217 splines in tension, Geophysics, 55, 293-305.
218GMT 4.5.6 10 Mar 2011 SURFACE(1)