Hi Vale!

Don't know if you have matlab but if you do

%% Open netcdf

ncid = netcdf.open('yourfile.nc','NC_NOWRITE');

%% Get all information about the netcdf file

fileinfo = ncinfo('yourfile.nc');

disp(fileinfo);

%% get info on a dimension

dim_0=netcdf.inqVar(ncid,0);

dim_1=netcdf.inqVar(ncid,1);

dim_X=netcdf.inqVar(ncid,X); (with X==max number dimension in the file)

%% get values corresponding to a dimension

var_0 = netcdf.getVar(ncid,0); (e.g. var_0=latitude)

var_1 = netcdf.getVar(ncid,1); (e.g. var_1=longitude)

var_X= netcdf.getVar(ncid,X); (e.g. var_X=SST)

All done!

Some ideas here vale: http://geog.uoregon.edu/bartlein/courses/geog607/Rmd/netCDF_01.htm Email me if you need help

Thank you Ana!

Hi Vale,

Have also used Ocean Data View to get an idea of what is in the file. https://odv.awi.de/

Also resources at:

http://www.esrl.noaa.gov/psd/data/gridded/tools.html

Mark

Hi Vale!

Don't know if you have matlab but if you do

%% Open netcdf

ncid = netcdf.open('yourfile.nc','NC_NOWRITE');

%% Get all information about the netcdf file

fileinfo = ncinfo('yourfile.nc');

disp(fileinfo);

%% get info on a dimension

dim_0=netcdf.inqVar(ncid,0);

dim_1=netcdf.inqVar(ncid,1);

dim_X=netcdf.inqVar(ncid,X); (with X==max number dimension in the file)

%% get values corresponding to a dimension

var_0 = netcdf.getVar(ncid,0); (e.g. var_0=latitude)

var_1 = netcdf.getVar(ncid,1); (e.g. var_1=longitude)

var_X= netcdf.getVar(ncid,X); (e.g. var_X=SST)

All done!

You can also do it using R, the netcdf package. I can send you some code if you need.

You can open netCDF files using 'esd' package in R. It is developed by R.Benestad ans AK Mezghani.

Thank you all  very much for the help!

For NATLAB, the free NCTOOLBOX makes reading a wide variety of data formats very easy.. https://github.com/nctoolbox/nctoolbox

For Python, the NetCFD4 module also provides easy access to single and multiple NetCDF datasets.  It's also free. https://pypi.python.org/pypi/netCDF4

You can use 'CDO' to select the particular locations and data levels you need for your study. I hope the CDO reference card attached proves useful. Email me if you need help.

The simplest way in UNIX or Linux would be just to use ncdump utility:

ncdump yourfile.nc > output.txt

That would generate an ascii file easy to read. Some examples can be found here:

http://www.unidata.ucar.edu/software/netcdf/workshops/2011/utilities/NcdumpExamples.html

Hope thsi may help.

Two more packages worth mentioning

Panoply, by NASA

http://www.giss.nasa.gov/tools/panoply/

UV-CDAT by LLNL

http://uvcdat.llnl.gov/index.html

Thank you all for your help. I managed to solve the problem using the RnetCDF package in R. Apparently on the MYOCEAN website there is a tool to suset their dataset but I have not found it yet.

Valentina Lauria, I can help you only in Matlab ... Can you work with that?

@Helle Ingelise Øvlis Jørgensbye

I did that manually for the WOA (NOAA oceanographic data).It comes as layers of oceanographic variables (e.g. salinity) in depth bands (0, 10, 20, 30, 50m … etc).

For each depth band, I made a raster of whether the bottom was in that specific band or not. Then I extracted the salinity values in the cells where the bottom was from the corresponding depth band salinity raster. That gave me a set of rasters (one per depth band) that I could mosaic together. (hope it is clear)

It did the job but it was a bit time consuming. I can send the script, if you know a bit about handling rasters in R, it should work

If there is an easier way to do that I would be happy to hear it though

 Here is the script, It will need some adaptation but hopefully it helps

Hi

You can extract data easily and convert to another format such as json, text, excel, csv with using Netcdf-Extractor

there is in  http://www.agrimetsoft.com/

best regard

Does any one knows how can I convert Netcdf format to hdf5? I would appreciate any help thar you can provide.

Its very simple only if you have ArcGIS Software.

Open Arcmap (Double click the Icon)

In the Arctoolbox-----Goto Multidimension Tools-----MakeNetCDFTableView (Click)

input netCDF file-------your file in '.nc' (NetCDF extension)

variable----------Precipitation or Temperature

Row Dimensions----------------Time stamps (Day or month)

Dimension Values-------------Lat and Long (You can give more values)

After the Process you can see your data and just export to excel or other statistical software.

Dear Isah;

Thank you very much for your answer. I think your proposed method works when we are interested in extracting statistical information from our data.

As I understood during this time, I found out netcdf can be introduced as the advanced version of hdf5.

However, thank you for your answer and information.

Best;

Sahar

Hi dear Valentina.

I think Panoply of NASA is a good choice as mentioned above, but the weakness is working with an individual point (one specific coordination) with every run.

more options discussed in this link:

https://www.researchgate.net/post/How_can_I_extract_a_time_series_of_variable_for_a_specific_location_specific_longitude_and_latitude_from_a_CMIP5_experiment_netCDF_data_set?discoverMore=1?view=5a33a8affb8931e31976d449

Hi Ahmad

Panoply is good for presenting data for all of the nc file. But it's not good for extracting data. Because you should read data of cells 'one-by-one'

and it's take many time or error(maybe). If you want extract data from one netcdf file you can use programming language(NCL, Matlab, R, Grads, Python, ... Or softwares as ArcGIS.)

But if you want to extract data from many files you should be a programmer.

If you need windows software for extracting easily, you can use Netcdf-Extractor in http://www.agrimetsoft.com/netcdf-extractor.aspx

Cheers

Dear Engineer Kolsoumi;

Thanks for your attention. Yes I have mentioned the weakness of Panoply above. Also I have downloaded your Netcdf-Extractor form your website, but as it was not free, I preferred to use other free softwares.

Kind Regards

I dowloaded Cordex RCPs for the precipitaion and temperature data for my study area, i want to use them for impact modeling. Does anyone know how to extract values from a netCDF?

Hi ahmadreza

You can visit this website for free softwares

https://www.unidata.ucar.edu/software/netcdf/software.html

Hi Edmud Mutayoba

Cordex Data has different domains. you can extract all domain files by netcdf-extractor V2. But you should calculate grid numbers. This work in 3 Domain is very difficult: SAM, AUS, ..?..

I'll develop a new software for all cordex domains that the user can enter coordinate and he/she doesn't need to calculate grid number.

Both of them are located in:

http://www.agrimetsoft.com/cordexDataExtractor.aspx

http://www.agrimetsoft.com/Netcdf-Extractor.aspx

What is the best way to do this?

For MyOcean data, you can use Motu-client to query a specific latitude/longitude/depth box. This will give you a much smaller netcdf file.

Then solutions mentioned in previous answers can be used to extract data. My favourite is python netCDF4 library. For example for CMEMS Mediterranean temperature dataset:

import netCDF4

ncf = netCDF4.Dataset('file.nc')

temperature = ncf['thetao'][:]

lat = ncf['lat'][:]

lon = ncf['lon'][:]

depth = ncf['depth'][:]

time = ncf['time'][:]

ncf.close()

Then you can do whatever you want.

Motu-client can be found here:

https://github.com/clstoulouse/motu-client-python/releases

Please view attachment.

Is this normal for I? If yes, how can I go about it to read the data.

Just for reference, in python you can use xarray:

import xarray as xr

ds = xr.open_dataset('name_of_a_netcdf_file.nc')

Here you will fing further info: http://xarray.pydata.org/en/stable/generated/xarray.open_dataset.html

An alternative in Python will be the IRIS library of the met-office. They have a nice tool that is quite intuitive, conceptually similar to the xarray library already commented by Raul Valenzuela

You only have to import the library and load the dataset, there are nice examples on the documentation of IRIS cubes on the web.

Beyond that you might be want to be familiar with the datasets first by exploring them.

ncdump -h netcdf_filename.nc

will give you and idea to the netcdf structure and

ncdump -v varname netcdf_filename.nc

will actually show the values on the terminal.

Take a look also to cdo as it has good summary properties

cdo -sinfon netcdf_filename.nc

will give you a nice overview of the dataset but explore all the operators on the cdo (climate-data-operators webpage).

Good luck!

hi. you can use panoply and gis ..

http://climatology.ir/?p=15935

Hello, I am quite new to NetCDF files, but I am using R and the ncdf4 package.

https://rpubs.com/boyerag/297592

This is for Python:

from netCDF4 import Dataset

a = Dataset('directory you have your trajectory file at')

b = data.variables.keys()

time = data.variables['time']

http://climatology.ir/?p=16050

use python, it's very simple, here is the example:

the code given below will open netcdf file including 3d variables such as var, lat, lon etc and after than it will save your data into excel/csv file.

install python 3.5+

########################################

import netCDF4

import xarray

import numpy as np

import csv

import pandas as pd

data = xarray.open_dataset('C:/NETCDFDATA/data.nc')

data.variables.keys() # select n run, it will show variables info

#reshape array 3d to 2d for saving var in csv/excel

var1 = data['variable'] # put name of the variable which you want to save

data1=np.resize(var1,[var1.shape[0],var1.shape[1]*var1.shape[2]])

data2=pd.DataFrame(data1)

data2.to_csv('C:/NETCDFDATA/data2.CSV')

#######################################

Hi

As @Dr.Vishal Singh has mentioned you can use Python. But for researchers and scholars that they don't know coding and writing a complete program, it wouldn't be easy, as well if there are huge number of files such as 30000 files or the files have daily scale. In this situation we need a tool with simple environment and also

we can easily work with it.

However, please visit the below link, then you can understand my mean.

https://agrimetsoft.com/netcdf-extractor.aspx

You know, when we want to extract a simple file we can do it with a simple code, but if we don't know coding and we want to extract data for a big number of files and implementing a code by creating loop and other criteria, It wouldn't be easy.

Cheers,

Nasrin

in matlab or Octave you can use:

ncdisp('your_file.nc')

%% to see what variables you have and then for example

lat= ncread('your_file.nc', 'latitude');

you can also use GIS Saga or ArcGIS following the steps: https://gis.ucar.edu/projects/working-netcdf-esri-arcgis

or R:

https://rpubs.com/boyerag/297592

i hope it helps,

naomi

Hi,

I know it is very late to respond but if some other one need the same, it will be very useful.

I am going to give you much simpler and lazy but effective way of doing this.

CDO will be very handy while working with netCDF files.

just inspect the levels in the file using ncdump -c filename.nc

choose which levels you want to include in the desired file and invoke the following CDO command..

cdo sellevel,1000,925,850,700,600,500,.... infile.nc outfile.nc

Cheers,

Kishore

If python, I recommend using xarray package.

http://xarray.pydata.org/en/stable/

Hello, script for data extraction from netcdf for particular location

library(chron)

library(RColorBrewer)

library(lattice)

library(ncdf.tools)

library(ncdf4)

library(RNetCDF)

setwd("~/ACCESS1-3")

ncin

U can use Panoply software to extract the data from netCDF file. Through Panoply you can also convert the netCDF file to .csv

Panoply is very simple and freely downloadable software

i know it is late to answer yiur question, but yiu can use GrADS

From what I understand the array is in z coordinates so the bottom level will be a different level for each location. So, only extracting a SINGLE LEVEL (as suggested a number of times) will not return the BOTTOM values at all grid points.

If this for only few arrays, it can be done in a loop in Matlab/Octave:

Assuming a M-by-N array t3d (with any number of levels), we get bottom field with

t3d=ncread('my_source_file.nc','variable_name'); % get 3D array

tb = NaN(M,N); % prepare array with bottom values

num = -100; % threshold value

for i=1:M,

for j=1:N,

k1=size(t3d(t3d(i,j,:) > num),3); % calculate number of levels containing values

if k1>0, tb(i,j)=t3d(i,j,k1); end % write into "tb" if you are not on land

end

end

[code to write tb into new netdcf file]

For this to work, values out of bounds must be NaN or below "num" and all regular values must be larger than "num" (as would be expected e.g. for temperature or salinity, otherwise, adapt "num").

However, if we have a large grid and/or large number of files, this will be quite time consuming.

Use GDAL software. And comand:

"gdal_translate –sds file_name.nc new_file .gtiff".

Or ArcGIS - ArcToolbox - Multidimensional data.

Hi Vignesh Prabhu

Yes Panoply is the best windows tool for extract data for a time cell but if you have a lot of time cell for example in daily scale of TRMM data if you want extract 30 years data so you will have more than 10950 cell in time dimension. So with Panoply you should change time dimension/copy/paste 10950 times and It's so high.

In result, for extract time series from netcdf files you should use Programming Language such as Matlab, NCl, NCO, R and stc.

Or if you don't want use the Programming Languages so you can use these two windows tool:

1- "Open NC File". Extract by using regular coordinates

2- "Netcdf-Extractor". Extract by using Grid Numbers

You can find then by search it in google.

Cheers

Write in scripts for the grads software:

'reinit'

'sdfopen C:\Data\hgt.nc'

'set lon 10 80'

'set lat 0 80'

'set gxout print'

'd hgt'

dummy=write('c:/Data/output.txt',result)

__________________

Write a loop to extract all the time