Trying to shift/recenter map and data

Hello, I have a 3-D dataset of sea surface temperatures. My matrix is of size 1440x721x1, where the first dimension is longitude, the second is lattitude, and the 3rd dimension is the averged sea surface temperatures at 1 time. My longtitude values run from 0 to 360 degrees, but I would like to recenter the data on the prime meridian (0 degrees). Ideally, I'd like to convert my 0 to 360 grid into a grid spanning from -180 to 180 and center the map and data on 180 degrees instead of 0. I've tried just about everything under the sun and end up with a result of either my data shifting and the map doesn't, or the map shifts and the data doesnt. Does anyone have any simple solutions to this?

figure(1)
pcolor(long,lat,array_1(:,:,1)');
shading flat
colorbar;xlabel('Longitude');ylabel('Latitude');
% axis([275 350 25 60]);
set(gca,'clim',[275 300]);

Answer 1

[Adam Danz]

You can wrap your longitudinal values to [-180,180] using wrapTo180() which requires the mapping toolbox.

Or you could use wrapToInterval() from the file exchange: wrapToInterval(long,[-180,180]);

Edit:

You'll also need the wrapped X data to be in the same order as the original X data and then resort the color value matrix (array_1) to match the new order of the wrapped X data. That can get complicated depending on how your x data are arranaged. Assuming the x is a vector in ascending order, you can follow this demo using the data you provided in a comment below. Note that data on x=0 and x=360 will be combined after wrapping.

D1 = load('long.mat');
long = D1.long;
D2 = load('lat.mat');
lat = D2.lat;
D3 = load('sst2.mat');
sst2 = D3.sst2; 

% Wrap longitudinal values and adjust order of C data according
% to the wrapped longitudinal values. This assumes the original
% longitudinal values are in ascending order.
assert(isequal(sort(long(:)), long(:)), 'This method assumes X was sorted in ascending order'); 
Xwrap = wrapTo180(long);
[XwrapSorted, XwrapOrder] = sort(Xwrap(:)); 
C = sst2(XwrapOrder,:);

% Plot original data
figure()
tiledlayout(2,1,'TileSpacing','compact')
nexttile()
pcolor(long,lat,sst2');
shading flat
colorbar;
xlabel('Longitude');
ylabel('Latitude');
set(gca,'clim',[275 300]);
title('Original data ([0,360])')
axis equal
axis tight

% Plot wrapped version
nexttile()
pcolor(XwrapSorted,lat,C');
shading flat
colorbar;xlabel('Longitude');ylabel('Latitude');
set(gca,'clim',[275 300]);
title('Wrapped data ([-180,180])')
axis equal
axis tight

[Brian DeCicco]

Hi Adam, thank you for your response! I tried your method of using: wrapTo180() for my longtitude file, but I am getting the same weird result where the longitude has adjusted, but my data does not (especially east of the prime meridian and over land). Please see image below and note how it differs from the initial plot I posted above. Is it possible that I need to wrap the data contained within my 3rd dimension some how as well?

[Adam Danz]

Good catch, but it will probably be easier to sort the wrapped X data than to rearrange the C data. Answer appended.

[Brian DeCicco]

Thanks Adam...I'll try and give that a shot and see what I get. In the mean time, I've just attached my data (.mat files) to better manipulate the data so you can see what I'm working with as well. ''sst2.mat'' is essentially the same as my "array_1(:,:,1)" matrix that I have in my original code when I posted.

[Brian DeCicco]

I tried your method above, but unfortunately the longitude values moved as they were supposed to, but my data did not, making it look off-kilter :( My data is attached in my last comment above. If you could take a look and see if you get the same result as I do that would be very much appreciated!

[Adam Danz]

Answer updated.

I'm curious, does 'sea surface temperatures' mean the floor surface or the water surface, I'm assuming the latter? Also, what units are the temperatures, Kelvin?

[Brian DeCicco]

Hey Adam! I gave it another shot and it works perfectly now. Looks like I had a minor typo and that was the reason why I couldn't get the data to shift over with the longitude grid using your code. It works perfectly now! Thank you so much for the quick and thorough responses, it is greatly appreciated. The temperatures in the dataset are in fact the water surface temperatures in degrees Kelvin.

-Brian

[JMSE]

Hi, I am facing a similar issue with the attached data. I want to recenter the data into a grid from 0 - 360. I am having a dataset with -180 to -180 and the aim is to extract data with lon: > 150 & 0 - 20. So, recentering the data would help to extract the desired dataset as one coherent set. With the snipped below I can extract a subset of e.g. lon > 150 & < 180, but not > 180 (or equivalent >0 within a 0-360 recentrered version.

So the example by @Adam Danz is already very helpful, but I am having a hard time to apply it to my case with the attached data.

Thank you!

dlat = load('prec_lat.mat');
lat = dlat.lat;
dlon = load('prec_lon.mat');
lon = dlon.lon;
dprec = load('prec.mat');
prec = dprec.prec;
J=find(lat > -30 & lat < 0);
K=find(lon > 150 & lon < 180);
precp=prec(J,K);
lonp=lon(K);
latp=lat(J);
figure;pcolor(lonp,latp,precp);shading flat
geoshow('landareas.shp', 'facecolor', 'k');
xlabel('longitude (°)'); ylabel('latitude (°)'); title('Precipitation 2021/12/19 00:00:00 - 00:00:29');
h=colorbar; 
set(get(h,'ylabel'),'String','Rain rate (mm/h)');
hold on
plot(178.19,-16,'r*')

Thanks!

Judith