1NCDIFF(1) General Commands Manual NCDIFF(1)
2
6 ncdiff - netCDF Differencer
7
9 ncdiff [-3] [-4] [-6] [-A] [-C] [-c] [-D dbg] [-d dim,[ min][,[ max]]]
10 [-F] [-h] [-L dfl_lvl] [-l path] [-O] [-p path] [-R] [-r] [-v
11 var[,...]] [-x] file_1 file_2 file_3
12
14 ncdiff subtracts variables in file_2 from the corresponding variables
15 (those with the same name) in file_1 and stores the results in file_3.
16 Variables in file_2 are broadcast to conform to the corresponding vari‐
17 able in file_1 if necessary. Broadcasting a variable means creating
18 data in non-existing dimensions from the data in existing dimensions.
19 For example, a two dimensional variable in file_2 can be subtracted
20 from a four, three, or two (but not one or zero) dimensional variable
21 (of the same name) in file_1. This functionality allows the user to
22 compute anomalies from the mean. Note that variables in file_1 are not
23 broadcast to conform to the dimensions in file_2. Thus, ncdiff, the
24 number of dimensions, or rank, of any processed variable in file_1 must
25 be greater than or equal to the rank of the same variable in file_2.
26 Furthermore, the size of all dimensions common to both file_1 and
27 file_2 must be equal.
28 When computing anomalies from the mean it is often the case that file_2
30 was created by applying an averaging operator to a file with the same
31 dimensions as file_1, if not file_1 itself. In these cases, creating
32 file_2 with ncra rather than ncwa will cause the ncdiff operation to
33 fail. For concreteness say the record dimension in file_1 is time. If
34 file_2 were created by averaging file_1 over the time dimension with
35 the ncra operator rather than with the ncwa operator, then file_2 will
36 have a time dimension of size 1 rather than having no time dimension at
37 all In this case the input files to ncdiff, file_1 and file_2, will
38 have unequally sized time dimensions which causes ncdiff to fail. To
39 prevent this from occuring, use ncwa to remove the time dimension from
40 file_2. An example is given below.
41 ncdiff will never difference coordinate variables or variables of type
43 NC_CHAR or NC_BYTE. This ensures that coordinates like (e.g., latitude
44 and longitude) are physically meaningful in the output file, file_3.
45 This behavior is hardcoded. ncdiff applies special rules to some NCAR
46 CSM fields (e.g., ORO). See NCAR CSM Conventions for a complete
47 description. Finally, we note that ncflint (ncflint netCDF File Inter‐
48 polator) can be also perform file subtraction (as well as addition,
49 multiplication and interpolation).
50
52 Say files 85_0112.nc and 86_0112.nc each contain 12 months of data.
53 Compute the change in the monthly averages from 1985 to 1986:
54 ncdiff 86_0112.nc 85_0112.nc 86m85_0112.nc
55 The following examples demonstrate the broadcasting feature of ncdiff.
57 Say we wish to compute the monthly anomalies of T from the yearly aver‐
58 age of T for the year 1985. First we create the 1985 average from the
59 monthly data, which is stored with the record dimension time.
60 ncra 85_0112.nc 85.nc
61 ncwa -O -a time 85.nc 85.nc
62 The second command, ncwa, gets rid of the time dimension of size 1 that
63 ncra left in 85.nc. Now none of the variables in 85.nc has a time
64 dimension. A quicker way to accomplish this is to use ncwa from the
65 beginning:
66 ncwa -a time 85_0112.nc 85.nc
67 We are now ready to use ncdiff to compute the anomalies for 1985:
68 ncdiff -v T 85_0112.nc 85.nc t_anm_85_0112.nc
69 Each of the 12 records in t_anm_85_0112.nc now contains the monthly
70 deviation of T from the annual mean of T for each gridpoint.
71 Say we wish to compute the monthly gridpoint anomalies from the zonal
73 annual mean. A zonal mean is a quantity that has been averaged over
74 the longitudinal (or x) direction. First we use ncwa to average over
75 longitudinal direction lon, creating xavg_85.nc, the zonal mean of
76 85.nc. Then we use ncdiff to subtract the zonal annual means from the
77 monthly gridpoint data:
78 ncwa -a lon 85.nc xavg_85.nc
79 ncdiff 85_0112.nc xavg_85.nc tx_anm_85_0112.nc
80 Assuming 85_0112.nc has dimensions time and lon, this example only
81 works if xavg_85.nc has no time or lon dimension.
82 As a final example, say we have five years of monthly data (i.e., 60
84 months) stored in 8501_8912.nc and we wish to create a file which con‐
85 tains the twelve month seasonal cycle of the average monthly anomaly
86 from the five-year mean of this data. The following method is just one
87 permutation of many which will accomplish the same result. First use
88 ncwa to create the file containing the five-year mean:
89 ncwa -a time 8501_8912.nc 8589.nc
90 Next use ncdiff to create a file containing the difference of each
91 month's data from the five-year mean:
92 ncdiff 8501_8912.nc 8589.nc t_anm_8501_8912.nc
93 Now use ncks to group the five January anomalies together in one file,
94 and use ncra to create the average anomaly for all five Januarys.
95 These commands are embedded in a shell loop so they are repeated for
96 all twelve months:
97 foreach idx (01 02 03 04 05 06 07 08 09 10 11 12)
98 ncks -F -d time,,,12 t_anm_8501_8912.nc foo.
99 ncra foo. t_anm_8589_.nc
100 end
101 Note that ncra understands the stride argument so the two commands
102 inside the loop may be combined into the single command
103 ncra -F -d time,,,12 t_anm_8501_8912.nc foo.
104 Finally, use ncrcat to concatenate the 12 average monthly anomaly files
105 into one twelve-record file which contains the entire seasonal cycle of
106 the monthly anomalies:
107 ncrcat t_anm_8589_??.nc t_anm_8589_0112.nc
108
111 NCO manual pages written by Charlie Zender and Brian Mays.
112
115 Report bugs to <http://sf.net/bugs/?group_id=3331>.
116
119 Copyright © 1995-2010 Charlie Zender
120 This is free software; see the source for copying conditions. There is
121 NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR
122 PURPOSE.
123
126 The full documentation for NCO is maintained as a Texinfo manual called
127 the NCO User's Guide. Because NCO is mathematical in nature, the docu‐
128 mentation includes TeX-intensive portions not viewable on character-
129 based displays. Hence the only complete and authoritative versions of
130 the NCO User's Guide are the PDF (recommended), DVI, and Postscript
131 versions at <http://nco.sf.net/nco.pdf>, <http://nco.sf.net/nco.dvi>,
132 and <http://nco.sf.net/nco.ps>, respectively. HTML and XML versions
133 are available at <http://nco.sf.net/nco.html> and
134 <http://nco.sf.net/nco.xml>, respectively.
135 If the info and NCO programs are properly installed at your site, the
137 command
138 info nco
140 should give you access to the complete manual, except for the TeX-
142 intensive portions.
143
146 The NCO homepage at <http://nco.sf.net> contains more information.
147 NCDIFF(1)