How to select the value of a variable given the latitude and the longitude in a NetCDF file?

21 March 2020 5 4K Report

Hi,

I'm new to this kind of data and I'm blocked.

I downloaded an analysis model for the CO from ENSEMBLE in NetCDF format (http://macc-raq-op.meteo.fr/index.php?category=ensemble&subensemble=hourly_ensemble&date=LAST&calculation-model=ENSEMBLE&species=o3&level=SFC&offset=000).

Using cdo (1.7.0) under ubuntu 16.04 I'm experiencing frustration.

The selection of the CO value using the remapnn interpolator returns identical values for many different latitudes. In the example below the same values are returned for all the latitudes between 40 and 49. Can someone specialist in the field explain to me what I'm doing wrong?

########## example

The name of the input .nc file W_fr -meteofrance,MODEL,ENSEMBLE+ANALYSIS+SURFACE+CO+-24H-1H_C_LFPW_20200320000000.nc is semplified to 20200220000000_co.nc

$ cdo -remapnn,lon=13.12_lat=40 20200220000000_co.nc $$.nc;cdo -outputtab,time,value,latitude,longitude $$.nc

cdo remapnn: Nearest neighbor weights from lonlat (700x400) to lonlat (1x1) grid

cdo remapnn: Processed 6720000 values from 1 variable over 24 timesteps ( 0.06s )

# time value lat lon

00:00:00 141.011 40 13.12

01:00:00 141.5 40 13.12

02:00:00 141.237 40 13.12

03:00:00 140.147 40 13.12

04:00:00 140.748 40 13.12

05:00:00 139.491 40 13.12

06:00:00 139.677 40 13.12

07:00:00 138.845 40 13.12

08:00:00 136.874 40 13.12

09:00:00 134.725 40 13.12

10:00:00 132.8 40 13.12

11:00:00 131.376 40 13.12

12:00:00 129.889 40 13.12

13:00:00 129.382 40 13.12

14:00:00 130.65 40 13.12

15:00:00 128.279 40 13.12

16:00:00 130.213 40 13.12

17:00:00 131.136 40 13.12

18:00:00 132.627 40 13.12

19:00:00 133.658 40 13.12

20:00:00 135.713 40 13.12

21:00:00 137.868 40 13.12

22:00:00 140.065 40 13.12

23:00:00 141.318 40 13.12

cdo outputtab: Processed 24 values from 1 variable over 24 timesteps ( 0.00s )

$ cdo -remapnn,lon=13.12_lat=49 20200220000000_co.nc $$.nc;cdo -outputtab,time,value,latitude,longitude $$.nc

cdo remapnn: Nearest neighbor weights from lonlat (700x400) to lonlat (1x1) grid

cdo remapnn: Processed 6720000 values from 1 variable over 24 timesteps ( 0.07s )

# time value lat lon

00:00:00 141.011 49 13.12

01:00:00 141.5 49 13.12

02:00:00 141.237 49 13.12

03:00:00 140.147 49 13.12

04:00:00 140.748 49 13.12

05:00:00 139.491 49 13.12

06:00:00 139.677 49 13.12

07:00:00 138.845 49 13.12

08:00:00 136.874 49 13.12

09:00:00 134.725 49 13.12

10:00:00 132.8 49 13.12

11:00:00 131.376 49 13.12

12:00:00 129.889 49 13.12

13:00:00 129.382 49 13.12

14:00:00 130.65 49 13.12

15:00:00 128.279 49 13.12

16:00:00 130.213 49 13.12

17:00:00 131.136 49 13.12

18:00:00 132.627 49 13.12

19:00:00 133.658 49 13.12

20:00:00 135.713 49 13.12

21:00:00 137.868 49 13.12

22:00:00 140.065 49 13.12

23:00:00 141.318 49 13.12

L. Augusto Sanabria

Hi Raffaele,

This is a recurrent question, see for instance these answers:

https://www.researchgate.net/post/How_can_I_extract_data_from_a_netCDF_file_for_a_specific_location#view=5e563a9e7ccd8270a430a915

Hope it helps,

Augusto

Raffaele Fronza

Hi Augusto,

I solved the issue using both raster+ncdf4 in R. Thank you for the help

Ismail Mondal

Step:

1. First Download Modis website label 1, 2 3 data

2. To covert NC file to Tiff format at ArcMap software

3. Then use the Arc Toolbox: Multidimension Tools- then make NetCDF to Raster layer add the .NC file…. then process

4. Then create a point shapefile to extract the IOP value

5. Extract Multi Values to points

6. Then export the value in Excel format

Narges Salehnia

Hi dear Fronza

Your question is so important. As you know read NetCDF file as nc or nc4 would be run with R, ArcGis, Matlab, NCL, etc., but however if you can develop a code for merging data, and extract them for a specific time series, it would not be easy. I invite you to see the bellow link. I have tried it, and it is really easy.

https://youtu.be/kR4Q59WQ6og

There are also similar answers that you can find them in RG, for example:

https://www.researchgate.net/post/Extracting_data_from_a_netCDF_file

Regards,

Narges

Thank you both for the answers.

I'm working in a Linux environment, where dynamically I can download, parse (100+ .nc files) and select (100+ regions) at a time; then the data are used to feed statistical and ML models. I favor using a strategy based on brutal scripting. I must use also R scripts as in bash I encountered issues in using `cdo`.

Need of infprmation about your Clean Sky Dissemination activities?

Who is involved on Abnormal neonatal eye movements?

How can Principal component analysis (or factor analysis) be applied on a questionnaire score, and how can the results be intepreted?

What is the criteria to be used to determine the elastic limit?

Is anyone aware of any relationship between alopecia and cardiovascular risk factors?

What kind of problem can you describe with non physical Hamiltonians?

Is there an equation of state capable to describe this type of copolymerization?

Are the autonomic seizures focal non motor as current ILAE classification propon?

Kinematics and Robot Design VI, KaRD2023

What is the reason for current dropping in OER , LSV curve?

Adhesion strength of coating?

Do interactions between biosphere, carbon cycle, & water cycle impact global warming & interaction between atmosphere & hydrosphere?

Who of all of you will help TPEOM to re-originate the Earths atmosphere to it original 5 atmospheric elements/gases?

How can I separate graphene from amorphous carbon of similar particle sizes?

Question about combustion Kinetics ?

What are the perfect mathematical model for a solar-powered ship implementation?

"How to handle large negative geopotential values at the surface in reanalysis datasets?

Why semicircle of nyquist plot is not complete?

What is the principle/mechanism behind aging of carbon (graphite) containing refractory mix for isostatically pressed refractories?