Reads in ERA-20C netCDF files and outputs a struct containing the
requested variables for a given spatial domain and time period.
ERA = get_ERA_forcing(Mobj, files, 'varlist', {'lon', 'lat', 'ssr'})
DESCRIPTION:
For the given netCDF(s), load the data into a struct for the variables
in varlist. If varlist is omitted, all variables are loaded.
INPUT:
Mobj - Mesh object containing the following fields:
* nativeCoords - 'spherical' or 'cartesian'
* lon, lat or x, y - node coordinates (depending on nativeCoords)
files - array of paths to the ERA-20C netCDF files
Optional keyword/value pairs:
'varlist' - give an array of variables to extract from the netCDF
files. Missing variables will cause an error.
'clip' - set to true for clipping the data to the model domain, false
to load the full data in the netCDFs.
OUTPUT:
era - struct containing the variables specified in varlist.
NOTES:
This script has been written against the netCDFs of the ECMWF-ERA20C
data created by the python script ecmwf-era20c.py.
Author(s)
Pierre Cazenave (Plymouth Marine Laboratory)
Revision history:
2015-08-13 First version based loosely on read_ERA_wind.m in the
fvcom-toolbox.
==========================================================================0001 function era = get_ERA_forcing(Mobj, files, varargin) 0002 % Reads in ERA-20C netCDF files and outputs a struct containing the 0003 % requested variables for a given spatial domain and time period. 0004 % 0005 % ERA = get_ERA_forcing(Mobj, files, 'varlist', {'lon', 'lat', 'ssr'}) 0006 % 0007 % DESCRIPTION: 0008 % For the given netCDF(s), load the data into a struct for the variables 0009 % in varlist. If varlist is omitted, all variables are loaded. 0010 % 0011 % INPUT: 0012 % Mobj - Mesh object containing the following fields: 0013 % * nativeCoords - 'spherical' or 'cartesian' 0014 % * lon, lat or x, y - node coordinates (depending on nativeCoords) 0015 % files - array of paths to the ERA-20C netCDF files 0016 % Optional keyword/value pairs: 0017 % 'varlist' - give an array of variables to extract from the netCDF 0018 % files. Missing variables will cause an error. 0019 % 'clip' - set to true for clipping the data to the model domain, false 0020 % to load the full data in the netCDFs. 0021 % 0022 % OUTPUT: 0023 % era - struct containing the variables specified in varlist. 0024 % 0025 % NOTES: 0026 % This script has been written against the netCDFs of the ECMWF-ERA20C 0027 % data created by the python script ecmwf-era20c.py. 0028 % 0029 % Author(s) 0030 % Pierre Cazenave (Plymouth Marine Laboratory) 0031 % 0032 % Revision history: 0033 % 2015-08-13 First version based loosely on read_ERA_wind.m in the 0034 % fvcom-toolbox. 0035 % 0036 %========================================================================== 0037 0038 subname = 'get_ERA_forcing'; 0039 0040 global ftbverbose 0041 if ftbverbose 0042 fprintf('\nbegin : %s \n', subname) 0043 end 0044 0045 assert(nargin >= 2, 'Incorrect number of input arguments (2 minimum).') 0046 0047 varlist = []; 0048 clip = false; 0049 for var = 1:2:length(varargin) 0050 switch varargin{var} 0051 case 'varlist' 0052 varlist = varargin{var + 1}; 0053 case 'clip' 0054 clip = varargin{var + 1}; 0055 end 0056 end 0057 0058 nf = length(files); 0059 0060 for f = 1:nf 0061 0062 assert(exist(files{f}, 'file') == 2, 'file: %s does not exist', files{f}) 0063 0064 if ftbverbose 0065 fprintf('Loading file %s (%i of %i)\n', files{f}, f, nf) 0066 end 0067 0068 if isempty(varlist) 0069 ncdetails = ncinfo(files{f}); 0070 varlist = {ncdetails.Variables.Name}; 0071 end 0072 0073 % Only do the spatial data on the first file (the data are global). 0074 if f == 1 0075 lon = double(ncread(files{f}, 'lon')); 0076 lat = double(ncread(files{f}, 'lat')); 0077 lonvector = unique(lon); 0078 latvector = unique(lat); 0079 dx = unique(diff(lonvector)); 0080 dy = mean(unique(diff(latvector))); 0081 0082 if clip 0083 % To save on memory, cull the data which doesn't come from the 0084 % domain we've specified (with a 2 element buffer). 0085 if min(Mobj.lon) < 0 && max(Mobj.lon) < 0 0086 % Easy, just shift by 360. 0087 idx_lon = find(lonvector > (min(Mobj.lon) - (2 * dx)) + 360 ... 0088 & lonvector < (max(Mobj.lon) + (2 * dx)) + 360); 0089 elseif min(Mobj.lon) < 0 && max(Mobj.lon) > 0 0090 % This is the tricky one. We'll do two passes to extract 0091 % the western chunk first (min(Mobj.lon) + 360 to 360), 0092 % then the eastern chunk (0 - max(Mobj.lon)) 0093 idx_lon{1} = find(lonvector >= (min(Mobj.lon) - (2 * dx)) + 360)'; 0094 idx_lon{2} = find(lonvector < (max(Mobj.lon) + (2 * dx)))'; 0095 idx_lon = [idx_lon{:}]; 0096 else 0097 % Dead easy, we're in the eastern hemisphere, so nothing 0098 % too strenuous here. 0099 idx_lon = find(lonvector > (min(Mobj.lon) - (2 * dx)) ... 0100 & lonvector < (max(Mobj.lon) + (2 * dx))); 0101 end 0102 % Latitudes are easy because there's only one system. 0103 idx_lat = find(latvector > (min(Mobj.lat) - (2 * dy)) & latvector < (max(Mobj.lat) + (2 * dy))); 0104 0105 else 0106 idx_lon = 1:length(lonvector); 0107 idx_lat = 1:length(latvector); 0108 end 0109 % Make a grid of the domain 0110 [era.lon, era.lat] = deal(lonvector(idx_lon), latvector(idx_lat)); 0111 era.lon(era.lon > 180) = era.lon(era.lon > 180) - 360; 0112 end 0113 0114 if ftbverbose 0115 fprintf('Variables:\n') 0116 end 0117 for v = 1:length(varlist) 0118 if ftbverbose 0119 fprintf('\t%s... ', varlist{v}) 0120 end 0121 vardump = double(ncread(files{f}, varlist{v})); 0122 era.(varlist{v}).data = vardump(idx_lon, idx_lat, :); 0123 era.(varlist{v}).lat = latvector(idx_lat); 0124 era.(varlist{v}).lon = lonvector(idx_lon); 0125 era.(varlist{v}).time = double(ncread(files{f}, 'time')); 0126 clear vardump 0127 if ftbverbose 0128 fprintf('done.\n') 0129 end 0130 end 0131 end 0132 0133 if ftbverbose 0134 fprintf('end : %s \n', subname) 0135 end