How to edit this code?

Hello all,

Recently I was trying to download ECMWF ERA5 precipitation data. the precipitation has 3 dimensions (lon x lat x time) the format of the file is NetCDF. data are in hourly time steps. each day begins at 00:00 and ends at 23:00. The NetCDF file includes hourly data for a year, so it has 365*24=8760 time steps. I would like to convert these hourly data daily but one main issue stands in front of me. I explain my issue with a very simple example:

to cover total precipitation for 1st January 2017 (for example), we need two days of data:

A) 1st January 2017 time = 01 - 23 will give you total precipitation data to cover 00 - 23 UTC for 1st January 2017

B) 2nd January 2017 time = 00 will give you total precipitation data to cover 23 - 24 UTC for 1st January 2017

So I need a code that Starts every day at 01:00 and ends at 00:00 next day and saves this value as a value of one day, rather than takes each day's time range between 00:00 to 23:00. And do it for each day in the year.

I have 2 types of code that unfortunately didn't consider this issue. if you kindly please help me edit this code I would be grateful. let me know if any other questions you have.

Code number 1:

filename='download.nc'; %Name of netcdf file
ncdisp(filename)       %show summary of nc
lat = ncread(filename,'latitude'); %reading latitude     
lon = ncread(filename,'longitude'); %reading longitude 
precip = ncread(filename,'tp'); %reading the main variable
t = ncread(filename,'time'); %read time

daily_precip = squeeze(nansum(reshape(precip, size(precip, 1), size(precip, 2), 24, []), 3));
dt = datetime(double(t)*3600, 'ConvertFrom', 'epochtime', 'Epoch', '01-Jan-1900');
[group, actualday] = discretize(dt, 'day');
daily_precip = splitapply(@(pages) {nansum(precip(:, :, pages), 3)}, (1:size(precip, 3))', group);
daily_precip = cat(3, daily_precip{:});

And the code numebr two:

filename='download.nc'; %Name of netcdf file
ncdisp(filename)       %show summary of nc
lat = ncread(filename,'latitude'); %reading latitude     
lon = ncread(filename,'longitude'); %reading longitude 
precip = ncread(filename,'tp'); %reading the main variable
t = ncread(filename,'time'); %read time

tp_daily=zeros(size(precip,1),size(precip,2),365);
for ii=0:364
    day=precip(:,:,ii*24+1:(ii+1)*24); %grab an entire day 
    tp_daily(:,:,ii+1)=nansum(day,3); % add the third dimension
end

please download my netcdf file using this link below:

https://drive.google.com/open?id=1qm5AGj5zRC3ifD1_V-ne2nDT1ch_Khik

Thank you everyone

[Mil Shastri]

It will be helpful if you can you attach your file directly to this thread. Also, I downloaded your file and it looks like its in a compiled (.nc) format. It will also be useful if you can attach the .csv or excel or text version of it.

[BN]

I was download it directly from ECWMF website. I haven't any other version of it :(

[Daniel M]

Note that length(0:23) is 24. So hourly data starting from 00:00 and going to 23:00 probably contains the full data for that day. I only think this because it would be of utmost nonsense to require a user to look at 2 days of data to capture data for 1 day.

[BN]

If the variable was something like temperature that records every hour and independent from later hours, this was true, but here the variable is total precipitation. Each hour precipitation of one past hour will be recorded. So if we want to know how much rainfall was accrued in a total of the day, we need to analyze 00:00 of next day to understand the value for 23:00 to 00:00. Actually, this data is for the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5, which recently was released.

In the ERA5 page, the python code was presented for this purpose.

https://confluence.ecmwf.int/display/CKB/ERA5%3A+How+to+calculate+daily+total+precipitation

But unfortunately, still, now, the Matlab code doesn't exist. It could be beneficial for all students and researchers around the world if any Matlab code for this purpose exists.

Thank You :)

[MarineM]

Hi, I have the same issue here. I am just wondering if 2 years later someone did discover a code for Matlab. Thank you :)