1 |
function [profOut]=MITprof_resample(profIn,fldIn,filOut,method); |
function [profOut]=MITprof_resample(profIn,fldIn,filOut,method); |
|
% |
|
2 |
%[profOut]=MITPROF_RESAMPLE(profIn,fldIn,filOut,method); |
%[profOut]=MITPROF_RESAMPLE(profIn,fldIn,filOut,method); |
3 |
% |
% |
4 |
% resamples a set of fields (fldIn) to profile 3D locations (profIn) |
% resamples a set of fields (specified in fldIn) to profile locations |
5 |
% and output the result to nc file (filOut) and memory (profOut). |
% (specified in profIn) and output the result either to memory |
6 |
% using a chosen interpolation method (method). |
% (by default) or to a netcdf file (if filOut is specified) based on |
7 |
% |
% on pre-defined interpolation method ('polygons' by default) |
8 |
% profIn (structure) should contain: prof_depth, prof_lon, prof_lat, prof_date |
% |
9 |
% fldIn (structure) should contain: fil, name, long_name, units, tim, |
% profIn (structure) should contain: prof_depth, prof_lon, prof_lat, |
10 |
% missing_value, FillValue (for nc output), and fld ([] by default). |
% and prof_date (serial date number from datenum.m) |
11 |
% If fld is not [] then it is assumed that user already loaded |
% |
12 |
% the fields from fldIn.fil. |
% fldIn (structure) should contain: fil, name, and tim (see below |
13 |
|
% for detail and examples), and optionally |
14 |
|
% - long_name, units, missing_value, FillValue; if filOut~='' |
15 |
|
% this information will be used in the netcdf fiel output |
16 |
|
% - fld ([] by default); if provided then it is assumed that |
17 |
|
% user has already read fldIn.fil and stored it to fldIn.fld |
18 |
% |
% |
19 |
% fldIn.tim can be set to |
% fldIn.tim must be set to one of the following values: |
20 |
% 'const' (for time invariant climatology), |
% 'const' (for time invariant climatology), |
21 |
% 'monclim' (for monthly climatology) |
% 'monclim' (for monthly climatology) |
22 |
% 'monser' (for monthly time series) |
% 'monser' (for monthly time series) |
23 |
% 'monloop' (for cyclic monthly time series) |
% 'monloop' (for cyclic monthly time series) |
24 |
% |
% |
25 |
% method can be set to |
% method ('polygons' by default) can be specified as |
26 |
% 'polygons' (linear in space) |
% 'polygons' (linear in space) |
27 |
% 'bindata' (nearest neighbor in space) |
% 'bindata' (nearest neighbor in space) |
28 |
% |
% |
29 |
% Example: |
% Example: (should be revisited) |
30 |
% grid_load; gcmfaces_global; MITprof_global; addpath matlab/; |
% |
31 |
% profIn=idma_float_plot('4900828'); |
% grid_load; gcmfaces_global; MITprof_global; addpath matlab/; |
32 |
% % |
% profIn=idma_float_plot('4900828'); |
33 |
% fldIn.fil=fullfile(myenv.MITprof_climdir,filesep,'T_OWPv1_M_eccollc_90x50.bin'); |
% % |
34 |
% fldIn.name='prof_Towp'; |
% fldIn.fil=fullfile(myenv.MITprof_climdir,filesep,'T_OWPv1_M_eccollc_90x50.bin'); |
35 |
% fldIn.long_name='pot. temp. estimate (OCCA-WOA-PHC combination)'; |
% fldIn.name='prof_Towp'; |
36 |
% fldIn.units='degree C'; |
% fldIn.tim='monclim'; |
37 |
% fldIn.tim='monclim'; |
% %fldIn.long_name='pot. temp. estimate (OCCA-WOA-PHC combination)'; |
38 |
% fldIn.missing_value=-9999.; |
% %fldIn.units='degree C'; |
39 |
% fldIn.FillValue=-9999.; |
% %fldIn.missing_value=-9999.; |
40 |
% fldIn.fld=[]; |
% %fldIn.FillValue=-9999.; |
41 |
% % |
% %fldIn.fld=[]; |
42 |
% profOut=MITprof_resample(profIn,fldIn); |
% % |
43 |
|
% profOut=MITprof_resample(profIn,fldIn); |
44 |
|
|
45 |
|
|
46 |
gcmfaces_global; |
gcmfaces_global; |
51 |
|
|
52 |
if isempty(who('method')); method='polygons'; end; |
if isempty(who('method')); method='polygons'; end; |
53 |
|
|
54 |
%0) check for file types |
%0) check for input types |
55 |
|
% test0=1 <-> binary |
56 |
|
% test1=1 <-> nctiles |
57 |
|
% test2=1 <-> readily available fldIn.fld |
58 |
test0=isfield(fldIn,'fil'); |
test0=isfield(fldIn,'fil'); |
59 |
|
% |
60 |
test1=0; |
test1=0; |
61 |
if test0; |
if test0; |
62 |
test0=~isempty(dir(fldIn.fil)); |
test0=~isempty(dir(fldIn.fil)); |
65 |
fil_nctiles=fullfile(PATH,NAME,NAME); |
fil_nctiles=fullfile(PATH,NAME,NAME); |
66 |
test1=~isempty(dir(fil_nc)); |
test1=~isempty(dir(fil_nc)); |
67 |
end; |
end; |
68 |
|
% |
69 |
|
if ~isfield(fldIn,'fld'); fldIn.fld=[]; end; |
70 |
test2=~isempty(fldIn.fld); |
test2=~isempty(fldIn.fld); |
71 |
|
|
72 |
%1) deal with time line |
%1) deal with time line |
73 |
if strcmp(fldIn.tim,'monclim'); |
if strcmp(fldIn.tim,'monclim'); |
74 |
tim_fld=[-0.5:12.5]; rec_fld=[12 1:12 1]; |
tmp1=[1:13]'; tmp2=ones(13,1)*[1991 1 1 0 0 0]; tmp2(:,2)=tmp1; |
75 |
|
tim_fld=datenum(tmp2)-datenum(1991,1,1); |
76 |
|
tim_fld=1/2*(tim_fld(1:12)+tim_fld(2:13)); |
77 |
|
tim_fld=[tim_fld(12)-365 tim_fld' tim_fld(1)+365]; rec_fld=[12 1:12 1]; |
78 |
|
% |
79 |
tmp1=datevec(profIn.prof_date); |
tmp1=datevec(profIn.prof_date); |
80 |
tmp2=datenum([tmp1(:,1) ones(profIn.np,2) zeros(profIn.np,3)]); |
tmp2=datenum([tmp1(:,1) ones(profIn.np,2) zeros(profIn.np,3)]); |
81 |
tim_prof=(profIn.prof_date-tmp2); |
tim_prof=(profIn.prof_date-tmp2); |
82 |
tim_prof(tim_prof>365)=365; |
tim_prof(tim_prof>365)=365; |
83 |
tim_prof=tim_prof/365*12;%neglecting differences in months length |
% |
84 |
if test2; fldIs3d=(length(size(fldIn.fld{1}))==4); end; |
if test2; fldIs3d=(length(size(fldIn.fld{1}))==4); end; |
85 |
elseif strcmp(fldIn.tim,'monloop')|strcmp(fldIn.tim,'monser'); |
elseif strcmp(fldIn.tim,'monloop')|strcmp(fldIn.tim,'monser'); |
86 |
if test1; |
if test1; |
87 |
eval(['ncload ' fil_nc ' tim']); |
eval(['ncload ' fil_nc ' tim']); |
88 |
nt=length(tim); |
nt=length(tim); |
89 |
elseif ~test2; |
elseif ~test2; |
90 |
|
warning('Here it is assumed that fldIn.fil contains 3D fields'); |
91 |
%note: 2D case still needs to be treated here ... or via fldIn.is3d ? |
%note: 2D case still needs to be treated here ... or via fldIn.is3d ? |
92 |
tmp1=dir(fldIn.fil); |
tmp1=dir(fldIn.fil); |
93 |
nt=tmp1.bytes/prod(mygrid.ioSize)/length(mygrid.RC)/4; |
nt=tmp1.bytes/prod(mygrid.ioSize)/length(mygrid.RC)/4; |
124 |
profOut=NaN*ones(profIn.np,profIn.nr); |
profOut=NaN*ones(profIn.np,profIn.nr); |
125 |
|
|
126 |
%2) loop over record pairs |
%2) loop over record pairs |
127 |
if ~strcmp(method,'bindata'); gcmfaces_bindata; end; |
if strcmp(method,'bindata'); gcmfaces_bindata; end; |
128 |
for tt=1:length(rec_fld)-1; |
for tt=1:length(rec_fld)-1; |
129 |
ii=find(tim_prof>=tim_fld(tt)&tim_prof<tim_fld(tt+1)); |
ii=find(tim_prof>=tim_fld(tt)&tim_prof<tim_fld(tt+1)); |
130 |
if ~isempty(ii); |
if ~isempty(ii); |
166 |
arr2=arr; |
arr2=arr; |
167 |
end; |
end; |
168 |
%now linear in time: |
%now linear in time: |
169 |
k0=floor(tim_prof(ii)); |
a0=(tim_prof(ii)-tim_fld(tt))/(tim_fld(tt+1)-tim_fld(tt)); |
|
a0=tim_prof(ii)-k0; |
|
170 |
if fldIs3d; |
if fldIs3d; |
171 |
a0=a0*ones(1,profIn.nr); |
a0=a0*ones(1,profIn.nr); |
172 |
profOut(ii,:)=(1-a0).*arr2(:,:,1)+a0.*arr2(:,:,2); |
profOut(ii,:)=(1-a0).*arr2(:,:,1)+a0.*arr2(:,:,2); |