How to extract data from .nc file based on latitude, longitude, time and wind?

Hello,

This is my first time trying to pull data from an netCDF file using MATLAB...

here's what I did...

clear all

close all

ncid = netcdf.open('uwnd.sig995.2016.nc','NC_NOWRITE');

lat = netcdf.getVar(ncid,0);

whos lat

lon = netcdf.getVar(ncid,1);

whos lon

time = netcdf.getVar(ncid,2);

whos time

wind = netcdf.getVar(ncid,3); %var3 wind 4x day

whos wind

latlim = [5 -5]; %5N 5S

lonlim = [-50 -45]; %-50W -45W

I dont know how to extrat one year for exemple 2016:

timelim = [ ? ]

and the wind for these lat lon and time...

Any help?

Thx in advance! Regards.

Below the data:

Format:
           netcdf4_classic

Global Attributes:
           Conventions   = 'COARDS'
           title         = '4x daily NMC reanalysis (2014)'
           history       = 'created 2013/12 by Hoop (netCDF2.3)'
           description   = 'Data is from NMC initialized reanalysis
                           (4x/day).  These are the 0.9950 sigma level values.'
           platform      = 'Model'
           References    = 'http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.html'
           dataset_title = 'NCEP-NCAR Reanalysis 1'

Dimensions:
           lat  = 73
           lon  = 144
           time = 1460  (UNLIMITED)

Variables:

      lat 
             Size:       73x1
             Dimensions: lat
             Datatype:   single
             Attributes:
                         units         = 'degrees_north'
                         actual_range  = [9.00e+01 -9.00e+01]
                         long_name     = 'Latitude'
                         standard_name = 'latitude'
                         axis          = 'Y'
      lon 
             Size:       144x1
             Dimensions: lon
             Datatype:   single
             Attributes:
                         units         = 'degrees_east'
                         long_name     = 'Longitude'
                         actual_range  = [0.00e+00 3.58e+02]
                         standard_name = 'longitude'
                         axis          = 'X'
      time
             Size:       1460x1
             Dimensions: time
             Datatype:   double
             Attributes:
                         long_name     = 'Time'
                         delta_t       = '0000-00-00 06:00:00'
                         standard_name = 'time'
                         axis          = 'T'
                         units         = 'hours since 1800-01-01 00:00:0.0'
                         actual_range  = [1.88e+06 1.89e+06]
      uwnd
             Size:       144x73x1460
             Dimensions: lon,lat,time
             Datatype:   single
             Attributes:
                         long_name               = '4xDaily u-wind at sigma level 995'
                         units                   = 'm/s'
                         precision               = 2
                         least_significant_digit = 1
                         GRIB_id                 = 33
                         GRIB_name               = 'UGRD'
                         var_desc                = 'u-wind'
                         level_desc              = 'Surface'
                         statistic               = 'Individual Obs'
                         parent_stat             = 'Other'
                         missing_value           = -9.97e+36
                         valid_range             = [-1.02e+02  1.02e+02]
                         dataset                 = 'NCEP Reanalysis'
                         actual_range            = [-3.45e+01  3.71e+01]

Answer 1

[Walter Roberson]

Your data is not arranged by year, it is arranged by time.

You will need to retrieve the Time vector. According to the notes it is in 'hours since 1800-01-01 00:00:0.0' . So divide that by 24 (to get days) and add datenum('1800-01-01 00:00:0') to get a result which is in MATLAB datenum format. Use datevec() to convert those values into arrays with year, month, day, hour, minute, second components. Compare to the desired year(s) to find the indices of the entries of interest. Then, use those indices to access the third dimension of the uwnd data in order to get the data for those times.

[Eliane Truccolo]

Hi Walter, Thank you! It was very easy!

[sara loren]

Hi, I have a similar question on hand. I have tried this code to extract the data, but I am unable to get the values, or link the values to lat,lon of the file. I want to get hourly values of precipitation (ppt) for each point. How do I get to link the lat lon with the values?

filesdir = 'source';
dir_info = dir( fullfile(filesdir, '*.nc') );
numberoffiles = length(dir_info);
filenames = fullfile( filesdir, {dir_info.name} );
ncvars =  {'lat','lon','ppt', 'time'};
lat = cell(numberoffiles, 1);
lon = cell(numberoffiles, 1);
ppt = cell(numberoffiles, 1);
times = cell(numberoffiles, 1);

for K = 1 : numberoffiles
  this_file = filenames{K};
  lat{K} = ncread(this_file, ncvars{1});
  lon{K} = ncread(this_file, ncvars{2});
  ppt{K} = ncread(this_file, ncvars{3});
  times{K} = ncread(this_file, ncvars{4});
end

[Walter Roberson]

We would need to see the ncfinfo for more detail, especially the units attribute.

[sara loren]

@Walter Roberson https://in.mathworks.com/matlabcentral/answers/740257-netcdf-data-extraction-linking-variables-to-time-lat-and-lon?s_tid=prof_contriblnk I have posted the question in detail here

Answer 2

[Rishwanth Darun A S]

hello Truccolo, Can you post the code for the same? Your revised code for the final result

[Walter Roberson]

from_year = 2014; to_year = 2015;   %for example

time = netcdf.getVar(ncid,2);
wind = netcdf.getVar(ncid,3); %var3 wind 4x day

time_datenum = time / 24 + datenum('1800-01-01 00:00:0');
dv = datevec(time_datenum);
date_match = dv(:,1) >= from_year & dv(:,1) <= to_year;

selected_wind = wind(:, :, date_match);

Or, expanding the selection,

from_date = '2014-04-01'; to_date = '2015-03-31';

time = netcdf.getVar(ncid,2);
wind = netcdf.getVar(ncid,3); %var3 wind 4x day

time_datenum = time / 24 + datenum('1800-01-01 00:00:0');
date_match = time_datenum >= datenum(from_date) && time_datenum <= datenum(to_date);

selected_wind = wind(:, :, date_match);

[Shilpa Hudnurkar]

thanks a lot.

[Malaidevan]

Now the nc file operates as no. of Hours from year 1900

So the corrected line for datevector will be:

"""

time_datenum = time / 24 + datenum('1900-01-01 00:00:00');

"""

[Walter Roberson]

Elaine's files happened to be based from 1800 not 1900.

nc files can use any base time they want, and do not have to use hours as the increment. We had a struggle a while back with a person whose nc file headers said that the time increment was 3 hours each, which failed badly when compared to what they expected the times to be, which was more like 20 minutes each.