please I need help to complete the descriptive statistics and representation of the following figures.

need help for the descriptive statistics of the following parameters depending on the year (time). but my variables are in 4D difficult for me to manipulate them please. my problem is with the extraction of carbonate variables, etc.. also the representation of these variables as a function of time in a format like (example: 2000,2003...)

need help for the descriptive statistics of the following parameters depending on the year (time). but my variables are in 4D difficult for me to manipulate them please. my problem is with the extraction of carbonate variables, etc.. also the representation of these variables as a function of time in a format like (example: 2000,2003...)

Ocolor=("global-analysis-forecast-bio-001-028-monthly_1693423231336.nc")
ncdisp (Ocolor)
lon=double(ncread(Ocolor,"longitude"));
lat=double(ncread(Ocolor,"latitude"));
dep=double(ncread(Ocolor,"depth")
[Lon,Lat]=meshgrid(lon,lat);
Time=double(ncread(Ocolor,"time"));
%Carbonate System
AT=ncread(Ocolor,"talk");
pbH=ncread(Ocolor,"ph");
DIC=ncread(Ocolor,"dissic");
PCO2=ncread(Ocolor,"spco2");
%Nutrient
NO3=ncread(Ocolor,"no3");
O2=ncread(Ocolor,"o2");
PO4=ncread(Ocolor,"po4");
Si=ncread(Ocolor,"si");
Fe=ncread(Ocolor,"fe");
%Primary Production
Phyc=ncread(Ocolor,"phyc"); %total phytoplankton
Nppv=ncread(Ocolor,"nppv"); %Total Primary Production of Phyto
Chl=ncread(Ocolor,"chl");

[dpb]

d=dir('*.zip');
unzip(d.name)
Ocolor=("global-analysis-forecast-bio-001-028-monthly_1693423231336.nc");
ncdisp (Ocolor);

OK, that looks pretty straightforward to index into by desired direction, give us an example of what specific "summary statistics" you're looking for for a given variable...

[Ebolo Nkongo]

@dpb Good evening please I would like to have the average, max, min, standard deviation. Linear regression between pH, talk, DIC, Chl for example. And represented for example the pH on a surface of longitude, latitude and depth. Then I could do the rest myself

Thank you in advance for your help.

Answer 1

[William Rose]

@Ebolo Nkongo,

fileIn="global-analysis-forecast-bio-001-028-monthly_1693423231336.nc";
t=ncread(fileIn,'time');

The time data is stored as integers representing hours since 01-Jan-1950. Therefore we convert hours to seconds and convert seconds to datetimes, as shown below:

dt=datetime(t*3600,'ConvertFrom','epochtime','Epoch','1950-01-01');

To confirm that the code above worked, we display the first three date-time values:

disp(dt(1:3)')

16-Nov-2020 00:00:00 16-Dec-2020 12:00:00 16-Jan-2021 12:00:00

Get info about the .nc file:

fInfo=ncinfo(fileIn);

Display fInfo.Dimensions (below). We see the data file includes data for 32 times, 8 latitudes, 9 longitudes, 3 depths.

Load some carbonate data and display the array size:

spco2=ncread(fileIn,"spco2");
disp(size(spco2))

9 8 32

This suggests that spco2 is given on a 9x8 long,lat grid, at 32 times.

You said you want descriptive statistics, and information across time. You can do that as follows:

spco2meanAll=mean(spco2,'All')          % mean spco2 across all locations and times

spco2meanAll = 36.2425

Next, we find the mean spCO2 at each time, by averaging across all locations. We do this by averaging across dimensions 1 and 2. We squeeze the result to eliminate uneeded dimensions from the result.

spco2meant=squeeze(mean(spco2,[1 2]));  % mean spco2 across all locations, each time is separate

Plot mean spco2 for each month versus time:

plot(dt,spco2meant,'-rx'); 
xlabel('Time'), ylabel('pCO_2 (Pa)')
grid on; title('Mean pCO_2 versus time')

which produces the figure below.

The plot shows that pCO2 reaches a local minimum in December of each year.

Best wishes with your research!

[Ebolo Nkongo]

Thank you very much this will help me. Would you have a course to help me improve please?

[William Rose]

@Ebolo Nkongo, you are welcome.

I'm afraid I do not have a course for improvement in Matlab. I never took a course in Matlab. I learned by trial and error, with careful reading of the help pages when I made mistakes - and I have made a lot of mistakes! Here are suggesitons for learning about Matlab. I have not checked the recommendations.

https://www.mathworks.com/matlabcentral/answers/8026-best-way-s-to-master-matlab

The question was posted a long time ago, but the answer has been updated more recently.

Good luck.

[Ebolo Nkongo]

@William Rose thank you .

please could you have a script that can allow me to have figures like the one I sent above? I would like to represent a distribution of this data. I used surf(lon,lat,pH) but it shows error.

[William Rose]

@Ebolo Nkongo,

unzip('global-analysis-forecast-bio-001-028-monthly_1693423231336.zip')
fileIn=("global-analysis-forecast-bio-001-028-monthly_1693423231336.nc");
fInfo=ncinfo(fileIn);
t=ncread(fileIn,'time');
dt=datetime(t*3600,'ConvertFrom','epochtime','Epoch','1950-01-01');
spco2=ncread(fileIn,"spco2");
latitude=ncread(fileIn,"latitude");
longitude=ncread(fileIn,"longitude");
spco2meanT=squeeze(mean(spco2,[1 2]));  % mean spco2 at each time, across all lat,long
spco2meanLL=mean(spco2,3);  % mean spco2 at each Lat & Long, across all times

Plot the mean pCO2 at each Lat,Long:

surf(latitude,longitude,spco2meanLL)
ylabel('Longitude (+E)'); xlabel('Latitude (+N)'); zlabel('pCO_2')
title('Time-averaged pCO_2 at each Lat,Long')
colorbar; view(90,-90)

This looks good.