Reads in ERA Interim files and outputs a struct containing the requested variables for a given year. ERA = read_ERA_wind(YEAR, DATADIR, VARLIST) DESCRIPTION: For the given YEAR, find all the ERA Interim NetCDF files and aggregate them into a single MATLAB struct, which contains the variables specified in VARLIST. In addition to the specified variables, time and longitude and latitude will also be extracted. INPUT: YEAR - year to extract DATADIR - path to the directory which contains the ERA NetCDF files VARLIST - list of the particular variables to extract from the NetCDF files. OUTPUT: era - struct containing the variables specified in VARLIST. Author(s) Pierre Cazenave (Plymouth Marine Laboratory) Revision history: 2012-10-19 First version based loosely on read_NCEP_wind.m in the fvcom-toolbox. ==========================================================================
0001 function era = read_ERA_wind(year, datadir, varlist) 0002 % Reads in ERA Interim files and outputs a struct containing the requested 0003 % variables for a given year. 0004 % 0005 % ERA = read_ERA_wind(YEAR, DATADIR, VARLIST) 0006 % 0007 % DESCRIPTION: 0008 % For the given YEAR, find all the ERA Interim NetCDF files and aggregate 0009 % them into a single MATLAB struct, which contains the variables 0010 % specified in VARLIST. In addition to the specified variables, time and 0011 % longitude and latitude will also be extracted. 0012 % 0013 % INPUT: 0014 % YEAR - year to extract 0015 % DATADIR - path to the directory which contains the ERA NetCDF files 0016 % VARLIST - list of the particular variables to extract from the NetCDF 0017 % files. 0018 % 0019 % OUTPUT: 0020 % era - struct containing the variables specified in VARLIST. 0021 % 0022 % Author(s) 0023 % Pierre Cazenave (Plymouth Marine Laboratory) 0024 % 0025 % Revision history: 0026 % 2012-10-19 First version based loosely on read_NCEP_wind.m in the 0027 % fvcom-toolbox. 0028 % 0029 %========================================================================== 0030 0031 % year = 2006; 0032 % % datadir = '/data/modellers/to_archive/momm-ERA40-interim/'; 0033 % datadir = '/users/modellers/pica/Data/ECMWF/2006'; 0034 % varlist = {'u10', 'v10'}; 0035 0036 warning off 0037 0038 if nargin ~= 3 0039 error('Incorrect number of arguments') 0040 end 0041 0042 subname = 'read_ERA_wind'; 0043 0044 global ftbverbose; 0045 if(ftbverbose); 0046 fprintf('\n') 0047 fprintf(['begin : ' subname '\n']) 0048 end; 0049 0050 if exist(datadir, 'dir') ~= 7 0051 error(['file: ' datadir ' does not exist']); 0052 end 0053 0054 %-------------------------------------------------------------------------- 0055 % Open ERA Interim data and check for time range 0056 %-------------------------------------------------------------------------- 0057 0058 % Get the time. 0059 ncERA = netcdf.open(fullfile(datadir, [num2str(year), '_U10V10.nc']), 'NOWRITE'); 0060 time_varid = netcdf.inqVarID(ncERA, 'time'); 0061 eratimehours = netcdf.getVar(ncERA, time_varid); 0062 0063 % ERA Interim times are stored as hours since 1900-01-01 00:00:0.0. 0064 % MATLAB's dates days since 0000/00/00 00:00:00. 0065 eratimehours = datevec((double(eratimehours)/24) + datenum('1900-01-01 00:00:0.0')); 0066 % Convert the ERA times to Modified Julian Date. 0067 era.time = greg2mjulian(eratimehours(:,1), eratimehours(:,2),... 0068 eratimehours(:,3), eratimehours(:,4), eratimehours(:,5),... 0069 eratimehours(:,6)); 0070 0071 if(ftbverbose); 0072 fprintf('beg time of ERA Interim data %04i/%02i/%02i %02i:%02i:%02i\n', eratimehours(1,:)); 0073 fprintf('end time of ERA Interim data %04i/%02i/%02i %02i:%02i:%02i\n', eratimehours(end,:)); 0074 end 0075 0076 % Get the geographical information from the ERA Interim data. Again, use 0077 % the U10 file only (we're assuming they're both global). 0078 lat_varid = netcdf.inqVarID(ncERA, 'latitude'); 0079 lon_varid = netcdf.inqVarID(ncERA, 'longitude'); 0080 eralatvector = netcdf.getVar(ncERA, lat_varid); 0081 eralonvector = netcdf.getVar(ncERA, lon_varid); 0082 [era.lon, era.lat] = meshgrid(eralonvector, eralatvector); 0083 0084 % Find the necessary variables 0085 for var=1:numel(varlist) 0086 0087 getVar = varlist{var}; 0088 varid_ERA = netcdf.inqVarID(ncERA, getVar); 0089 0090 % Get the data 0091 data = netcdf.getVar(ncERA, varid_ERA, 'single'); 0092 0093 if strcmpi(getVar, 'u10') || strcmpi(getVar, 'v10') 0094 % The ERA Interim wind component data are packed as integers. The 0095 % following equation describes how to unpack them: 0096 % unpacked value = add_offset + ((packed value)*scale_factor) 0097 % (from 0098 % http://www.ecmwf.int/products/data/archive/data_faq.html#netcdfintegers). 0099 % ERA wind scale_factor and add_offset are doubles (the NCEP ones 0100 % are singles). 0101 scale_factor = netcdf.getAtt(ncERA,varid_ERA,'scale_factor','double'); 0102 add_offset = netcdf.getAtt(ncERA,varid_ERA,'add_offset','double'); 0103 0104 % Unpack the values. In general, the data for U10 and V10 should be 0105 % doubles for griddata to work. Fix the order of the dimensions to 0106 % match the coordinates in eralon and eralat. 0107 era.(getVar) = permute(double(add_offset + (data.*scale_factor)), [2,1,3]); 0108 else 0109 % We're assuming they're not packed and so we just return the data 0110 % as is (but as doubles). 0111 era.(getVar) = double(data); 0112 end 0113 end 0114 0115 netcdf.close(ncERA) 0116 0117 if ftbverbose; 0118 fprintf(['end : ' subname '\n']) 0119 end