Analyse Time Series Data based on Seasonality

Hello everyone,

I'm trying to analyse seasonal dataset from 2015 to 2022. I have total 88 'nc' file. file name as "H08_20190301_0000_1MSST200_FLDK.06001_06001.ncI". I aims to seperate data into four seasons (e.g., Spring from March to May). I used coding below to store all data into cell structure.

My problem is I don't know how to scanning the file name to seperate and integrate data into season. For example, this data name ""H08_20190301_0000_1MSST200_FLDK.06001_06001.ncI" provides time in 20190301 belonging to Spring. I need to read specific date time to store all data with same season into one strucuture. Please help me to deal with this. Thank in advance!

pth = 'C:\Users\LongHoang\OneDrive\Desktop\Spring_2023\Ocean_Atmosphere_Mathlab\Mid_Term\H8_SST\H8_SST\';
ncvars =  {'lat', 'lon', 'sea_surface_temperature'};

dinfo = dir( fullfile(pth, '*.nc') );
num_files = length(dinfo);
filenames = fullfile( pth, {dinfo.name} );

name = fullfile({dinfo.name}); %I have cell file to store all file name but don't know how to scan the file to get 
% date time

lats = cell(num_files, 1);
lons = cell(num_files, 1);
ssts = cell(num_files, 1);

for k = 1: num_files
  
  this_file = filenames{k};
  Name_file = name{k};
  lats{k} = ncread(this_file, ncvars{1});
  lons{k} = ncread(this_file, ncvars{2});
  ssts{k} = ncread(this_file, ncvars{3});

end

[Peter Perkins]

Long, it's not clear what are in your files. Does each file contain one lat-by-lon SST array for each season in each year? That sounds wrong, because that would be 8*4=32 files, not 88. Does each file contain daily SST values, i.e. lat-by-lon-by-day, over some range, like monthly files?

Answer 1

[Mathieu NOE]

hello

I wished I had some files to test the code , but let's go directly to the task

I simply extracted from the filename the number corresponding to the month , then you make a simple test to store the data in the correct array depending of the month number

I was lazy so I created explicitely one array per season like lats_spring , lats_summer, lats_fall , lats_winter

but you could change that to a structure or cell array if you prefer

also for each season instead or storing in separate cells you might want simply to concatenate the data and have one unique array at the end (your preference ?)

could be something like this :

pth = 'C:\Users\LongHoang\OneDrive\Desktop\Spring_2023\Ocean_Atmosphere_Mathlab\Mid_Term\H8_SST\H8_SST\';
ncvars =  {'lat', 'lon', 'sea_surface_temperature'};
dinfo = dir( fullfile(pth, '*.nc') );
num_files = length(dinfo);
filenames = fullfile( pth, {dinfo.name} );
name = fullfile({dinfo.name}); %I have cell file to store all file name but don't know how to scan the file to get 
% date time
lats = cell(num_files, 1);
lons = cell(num_files, 1);
ssts = cell(num_files, 1);

k1 = 0; % "winter" counter init
k2 = 0; % "spring" counter init
k3 = 0; % "summer" counter init
k4 = 0; % "fall" counter init

for k = 1: num_files
  
  this_file = filenames{k};
  Name_file = name{k};
    
    % new code
    fulldate = split(Name_file,'_');
    fulldate = fulldate{2}; % assuming the date appears always in second position in the filename (between '_')
    monthsnumber = str2num(fulldate(5:6)); % to have the month corresponding to the file

    if monthsnumber >= 12 && monthsnumber <= 2 % winter
       k1 = k1 + 1; % counter
      lats_winter{k1} = ncread(this_file, ncvars{1});
      lons_winter{k1} = ncread(this_file, ncvars{2});
      ssts_winter{k1} = ncread(this_file, ncvars{3});
    end
    
    if monthsnumber >= 3 && monthsnumber <= 5 % spring
       k2 = k2 + 1; % counter
      lats_spring{k2} = ncread(this_file, ncvars{1});
      lons_spring{k2} = ncread(this_file, ncvars{2});
      ssts_spring{k2} = ncread(this_file, ncvars{3});
    end

    if monthsnumber >= 6 && monthsnumber <= 8 % summer
       k3 = k3 + 1; % counter
      lats_summer{k3} = ncread(this_file, ncvars{1});
      lons_summer{k3} = ncread(this_file, ncvars{2});
      ssts_summer{k3} = ncread(this_file, ncvars{3});
    end

    if monthsnumber >= 9 && monthsnumber <= 11 % fall
       k4 = k4 + 1; % counter
      lats_fall{k4} = ncread(this_file, ncvars{1});
      lons_fall{k4} = ncread(this_file, ncvars{2});
      ssts_fall{k4} = ncread(this_file, ncvars{3});
    end
end

[Mathieu NOE]

hello

problem solved ?

[Long]

Sorry for late reply. I had some personal things to deal. It work perfectly now. Thank you lots !

[Mathieu NOE]

would you mind accepting my answer ?

thanks

[Mathieu NOE]

would you mind accepting my answer ?

thanks

[Mathieu NOE]

Hello

Problem solved ?

would you mind accepting my answer ? thanks !

[Long]

I had accepted your answer. Sorry for some inconvience stuff.