How to convert an ASCII to NetCDF?

Hi, i'm trying to convert a large ASCII file into a NetCDF through matlab. It's a climate file 15329 x 16 columns.

nccreate('monthly_tmp.nc', 'monthly_temperatures', 'Dimensions', {'ff' 12 'lon' 180 'lat' 180}, 'Format','classic')
fid=fopen('tmp.mat') 
var=zeros(12)
while ~feof(fid) 
   coor=fscanf(fid, '%d', 2);
   var=fscanf(fid, '%f', inf); 
   lat=coor(1); lon=coor(2);
   ncwrite('monthly_tmp.nc', 'monthly_temperatures', var (1 lat, lon)) 
end

The columns are set out as latitude, longitude, year and 12 monthly values.

I need to first read lat and lon and then assign a monthly value to that grid.

So i need columns 1 and 2, and then 4-15 for my temperature values.

I'm really stuck with how I can feed my variables into the loop!

Thanks for your help!

Answer 1

[Ashish Uthama]

Hope this helps you get started:

%%set up dummy data
fid = fopen('tmp.mat','w');
fmtStr = repmat(' %f',[1,13]);
for lonInd=1:18
    for latInd=1:18
        fprintf(fid,['%d %d' fmtStr,'\n'], lonInd, latInd, 2013,(1:12)*latInd*lonInd);
    end
end
fclose(fid);
nccreate('monthly_tmp.nc', 'monthly_temperatures', 'Dimensions', {'ff' 12 'lon' 18 'lat' 18}, 'Format','classic')
fid=fopen('tmp.mat');
while ~feof(fid)
   coor=fscanf(fid, '%d', 2);
   var =fscanf(fid, '%f', 13);
   % assuming the lat and lon are whole numbers from 1 to 180
   lat=coor(1); lon=coor(2);
   % assuming var(1) is the year
   temp = var(2:end);
   ncwrite('monthly_tmp.nc', 'monthly_temperatures', temp, [1 lon lat]);
end
fclose(fid);

%%spot check
d = ncread('monthly_tmp.nc','monthly_temperatures',[1,2,2],[12, 1,1])

Answer 2

[Darwin Aramburo Palacios]

Good morning I have several series of wave spectrum (Matrix of 110 rows * 121 columns) in .txt files, the data is configured in the following dates

Years: 2015 to 2018

Month: January to December

Day: 1:31

Time from 00 to 23

My intention is to convert these files from .txt format to netcdf format for YEARS

Example of how the data is written

SpecDimarAno_2012_Mes_1_Dia_1_Hora_0

SpecDimarAno_2012_Mes_1_Dia_2_Hora_1

.

.

.

I appreciate if you can give me advice

[Walter Roberson]

110 * 121 / 365 / 24

ans = 1.5194

Your file appears to have enough information for 36 readings per day instead of 24 readings per day, so it is not obvious how the data is organized? The organization will be important for creating proper netcdf indices.

[Darwin Aramburo Palacios]

actually

SpecDimarAno_2012_Mes_1_Dia_1_Hora_0

SpecDimarAno_2012_Mes_1_Dia_1_Hora_1

SpecDimarAno_2012_Mes_1_Dia_1_Hora_2

SpecDimarAno_2012_Mes_1_Dia_1_Hora_3

.

.

.

SpecDimarAno_2012_Mes_1_Dia_1_Hora_23

so it continues for the other months and other years