-
Notifications
You must be signed in to change notification settings - Fork 6
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #1 from ArgoDMQC/develop
Develop
- Loading branch information
Showing
14 changed files
with
399 additions
and
51 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,2 +1,4 @@ | ||
| ow_config_* | ||
| ow_calibration_* | ||
| set_calseries_* | ||
| ~$* |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,2 +1,63 @@ | ||
| # matlabow | ||
| The OW Matlab toolbox | ||
| # The Matlab OW toolbox | ||
|
|
||
| This is a package of MATLAB routines for calibrating profiling float conductivity sensor drift. A description of the algorithms can be found in "An improved calibration method for the drift of the conductivity sensor on autonomous CTD profiling floats by θ-S climatology", by W.B. Owens and A.P.S. Wong, in Deep-Sea Research Part I: Oceanographic Research Papers, 56(3), 450-457, 2009. | ||
| Lately, modifications suggested in “Improvement of bias detection in Argo float conductivity sensors and its application in the North Atlantic” , by C. Cabanes, V. Thierry and C. Lagadec, in Deep-Sea Research Part I, 114, 128-136, 2016 have been taken into account. | ||
|
|
||
| # How to install the toolbox ? | ||
|
|
||
| Either clone the latest version of the git repository: | ||
|
|
||
| git clone https://github.com/ArgoDMQC/matlabow.git | ||
|
|
||
| or download and unzip the zip file (Clone or download button) | ||
|
|
||
| You can also access the different releases here: | ||
| https://github.com/ArgoDMQC/matlabow/releases | ||
|
|
||
| # How to run the analysis? | ||
| Here is a summary of what should be done to run the analysis, please read the ./doc/README.doc file for more details | ||
|
|
||
|
|
||
| 1. All files are to be used in MATLAB. The full package was tested with MATLAB R2014a. In addition, you will need: | ||
| a). The MATLAB Optimization Toolbox; | ||
| b). The ITS-90 version of the CSIRO SEAWATER library. The version 3\_3.0 of this library can be found in ./lib/seawater\_330\_its90. Please update if necessary. | ||
| c). The M_MAP toolbox. The version 1.4.c of this library can be found in ./lib/m\_map1.4. Please update if necessary. | ||
|
|
||
| 2. Add the necessary path to your matlab path: addpath('./lib/seawater\_330\_its90';'./lib/m\_map1.4';'./matlab\_codes/') | ||
|
|
||
| 3. Put your reference data in ./data/climatology/historical\_ctd, /historical\_bot, /argo\_profiles. | ||
|
|
||
| REFERENCE DATA can be obtain at ftp.ifremer.fr | ||
| cd /coriolis/data/DMQC-ARGO/ (if you need a login/pswd ask codac@ifremer.fr) | ||
|
|
||
| Then, create/update your ./data/constants/wmo\_boxes.mat file (more details in ./doc/README.doc, p3) | ||
|
|
||
| 4. After you have decided where you want to install the package on your computer, edit ow\_config.txt at the following lines so the correct pathways are specified: | ||
|
|
||
| * HISTORICAL\_DIRECTORY = | ||
| * FLOAT\_SOURCE\_DIRECTORY = | ||
| * FLOAT\_MAPPED\_DIRECTORY = | ||
| * FLOAT\_CALIB\_DIRECTORY = | ||
| * FLOAT\_PLOTS\_DIRECTORY = | ||
| * CONFIG\_DIRECTORY = | ||
|
|
||
| 5. The last section of ow\_config.txt below the heading "Objective Mapping Parameters" is where you set the various parameters (more details in .doc.README.doc, p4-6) | ||
|
|
||
| 6. If this is the first time you are using this system, then the 4 directories /data/float\_source, /float\_mapped, /float\_calib, and /float\_plots should be empty. Decide how you want to organise your floats, e.g. under different project names or different investigator names. Then make identical subdirectories under each of these 4 directories. For example: | ||
|
|
||
| /data/float\_source/project\_xx | ||
| /data/float\_mapped/project\_xx | ||
| /data/float\_calib/project\_xx | ||
| /data/float\_plots/project\_xx | ||
|
|
||
| 7. Create the float source file (./data/float\_source/project\_xx/$flt\_name$.mat) from the original netcdf files (more details in ./doc/README.doc,p6) | ||
|
|
||
| 8. Open MATLAB in the top directory. List all the float files in a cell array "float\_names", with the corresponding subdirectories in another cell array "float_dirs". For example, | ||
| float_dirs = { 'project\_xx/'; 'project\_xx/'; 'jones/'; 'jones/' }; | ||
| float_names = { 'float0001'; 'float0002'; 'myfloat\_a'; 'myfloat\_b' }. | ||
|
|
||
| Tips: If the files are not saved under a subdirectory and are only saved under ./float\_source/, specify float\_dirs = { ''; ''; ''; '' }, etc. | ||
|
|
||
| Run ow\_calibration.m. | ||
|
|
||
|
|
Binary file not shown.
Binary file not shown.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,142 @@ | ||
|
|
||
| function cov = build_cov( ptmp,coord_float,po_system_configuration); | ||
|
|
||
| % This function was created from build_ptmp_cov which the description is written below: | ||
| % | ||
| % Description of build_ptmp_cov | ||
| % ____ | ||
| % This function builds the covariance matrix (a square matrix) that has nxn number | ||
| % of tiles, each tile is of mxm size. | ||
| % | ||
| % The vertical covariance matrix is the building tile. It has 1 at its diagonal, | ||
| % then decreases exponentially in the off-diagonals, which represents the vertical | ||
| % covariance between water masses. | ||
| % | ||
| % This version assums that each profile is independent from the other ones. | ||
| % | ||
| % Annie Wong, May 2001 | ||
| % Breck Owens, November 2007 | ||
| %____ | ||
| % | ||
| % Cecile Cabanes, June 2013 | ||
| % Compared to build_ptmp_cov, build_cov( ptmp,coord_float,po_system_configuration) take into account the horizontal covariance between different mapped profiles. | ||
| % This lateral covariance take into account the fact that a mapped profile on a Argo position is build from a set of historical profiles that is not very different from the set used to build a mapped profile at the next or previous Argo profile position. | ||
| % This lateral covariance between two mapped profiles is constructed using a Gaussian function and the large spatial scales | ||
| % | ||
| % Cecile Cabanes, 2017 | ||
| % use the small spatial scales instead of the large spatial scales to build | ||
| % the lateral covariance: found to be the best compromise between | ||
| % informative errors and large enough NDF for AIC criterium, at least for | ||
| % the scales defined for the North Atlantic bassin | ||
|
|
||
| % Set up the theta boundaries for water masses | ||
|
|
||
| ptboundaries = [ 30; 24; 18; 12; 8; 4; 2.5; 1; -2 ]; | ||
| ptscale_down = [ 6; 6; 6; 4; 4; 1.5; 1.5; 1; 1 ]; | ||
| ptscale_up = [ 6; 6; 6; 6; 4; 4; 1.5; 1.5; 1 ]; | ||
|
|
||
|
|
||
| % set up the building tile = vertical covariance matrix | ||
| % | ||
| % upper triangle of the matrix = covariance of each ptlevel with every ptlevel below it, | ||
| % looking down the water column from the diagonal | ||
| % lower triangle of the matrix = covariance of each ptlevel with every ptlevel above it, | ||
| % looking up the water column from the diagonal | ||
|
|
||
| [m,n]=size(ptmp); | ||
|
|
||
| cov=zeros(m*n,m); % set up the output covariance matrix, with 1's along the diagonal | ||
|
|
||
|
|
||
| for k=1:n % for each profile | ||
| k1 = (k-1)*m; | ||
| for l=1:1 % for each profile | ||
| l1 = (l-1)*m; | ||
| for i=1:m % for all levels | ||
| for j=1:m | ||
| if(i<j) % upper triangle, look down the water column for vertical scale | ||
| Ltheta = interp1( ptboundaries, ptscale_down, ptmp(i,k), 'linear' ); | ||
| cov(i+k1,j+l1) = exp( - ( ptmp(j,k) - ptmp(i,k) ).^2/ Ltheta.^2 ); | ||
| elseif(i>j) % lower triangle, look up the water column for vertical scale | ||
| Ltheta = interp1( ptboundaries, ptscale_up, ptmp(i,k), 'linear' ); | ||
| cov(i+k1,j+l1) = exp( - ( ptmp(j,k) - ptmp(i,k) ).^2/ Ltheta.^2 ); | ||
| elseif(i==j) | ||
| cov(i+k1,j+l1) =1; | ||
| end | ||
| if isnan(cov(i+k1,j+l1)) | ||
| cov(i+k1,j+l1)=1; | ||
| end | ||
| end | ||
| end | ||
| end | ||
| end | ||
|
|
||
|
|
||
| % | ||
| % set up the horizontal covariance matrix | ||
| scale_lon = str2num(po_system_configuration.MAPSCALE_LONGITUDE_SMALL); | ||
| scale_lat = str2num(po_system_configuration.MAPSCALE_LATITUDE_SMALL); | ||
| scale_phi = str2num(po_system_configuration.MAPSCALE_PHI_SMALL); | ||
| map_use_pv = str2num(po_system_configuration.MAP_USE_PV); | ||
|
|
||
| for k=1:n | ||
| b(k,:)=covarxy_pv(coord_float(k,:),coord_float,scale_lon,scale_lat,scale_phi,map_use_pv); | ||
| end | ||
| b=b(:,1:n); | ||
|
|
||
| covh=b; | ||
|
|
||
| %keyboard | ||
| % build the final covariance matrix, taking into account horizontal and vertical covariance | ||
|
|
||
| covn=repmat(cov,[1,n]); | ||
| for k=1:n % for each profile | ||
| kd = (k-1)*m+1; | ||
| kf = k*m; | ||
| for l=1:n % for each profile | ||
| ld = (l-1)*m+1; | ||
| lf = (l)*m; | ||
| covn(kd:kf,ld:lf)=covh(k,l)*covn(kd:kf,ld:lf); | ||
| end | ||
| end | ||
|
|
||
| cov=covn; | ||
|
|
||
| return | ||
|
|
||
| %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% | ||
|
|
||
| % function [a] = covarxy_pv(x1,x2,gs,ts,phi,pv) | ||
|
|
||
| function [a] = covarxy_pv(x1,x2,gs,ts, phi, pv) | ||
|
|
||
| m=size(x1); | ||
| n=size(x2); | ||
| a=zeros(m(1),n(1)); | ||
| one=ones(n(1),1); | ||
|
|
||
|
|
||
| % derive planetary vorticity at each data point ---- | ||
|
|
||
| Zx1=x1(:,3); % depth at x1 datapoint | ||
| Zx2=x2(:,3); % depth at x2 datapoint | ||
|
|
||
| PV_x1=(2*7.292*10^-5.*sin(x1(:,1).*pi/180))./Zx1; | ||
| PV_x2=(2*7.292*10^-5.*sin(x2(:,1).*pi/180))./Zx2; | ||
|
|
||
|
|
||
| % calculate covariance term, use pv as optional ---- | ||
|
|
||
| for i=1:m(1) | ||
| tmp=((x1(i,1)-x2(:,1))./ts).^2 + ... | ||
| ((x1(i,2)-x2(:,2))./gs).^2; | ||
| if(pv &PV_x1~=0&PV_x2~=0) % if PV is wanted and the point is not on the equator --- | ||
| tmp = tmp + ... | ||
| ( (PV_x1(i)-PV_x2)./sqrt( PV_x1(i).^2+PV_x2.^2 )./phi ).^2; | ||
| end | ||
|
|
||
|
|
||
| a(i,:)=exp(-tmp'); | ||
| end | ||
|
|
||
| return |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.