It would help if you provided some more specifics... are your sensors arranged in a grid-like pattern, such that it makes sense to show the heatmap in a pcolor-like way, or are colored points, like a scatter plot, more appropriate?
An example of the latter is below (I arbitrarily arranged the sensors in a circle). The only difference between doing an animation on a projected axis vs a cartesian axis is that the plot object (in this example, the scatter object hs) holds post-projected x/y coordinates rather than the lat/lon coordinates used as input to scatterm, so you have to replicate this projection calculation prior to updating the scatter object x/y data.
% Center of array, at 100 time steps
clt = linspace(-80, 80, 100);
cln = linspace(0, 360, 100);
% Value of sensor
val = rand(18,100);
% Sensor location with respect to center
r = km2deg(1000);
th = linspace(0,2*pi,19);
th = th(1:end-1)';
dlt = r .* cos(th);
dln = r .* sin(th);
% Sensor location
lt = bsxfun(@plus, clt, dlt);
ln = wrapTo360(bsxfun(@plus, cln, dln));
% Plot time step 1
hax = worldmap('world');
C = load('coast');
geoshow(C);
hs = scatterm(lt(:,1), ln(:,1), [], val(:,1), 'filled');
% Update x-, y-, and cdata to produce animation
[x,y] = mfwdtran(lt, ln);
for it = 2:size(val,2)
set(hs.Children, 'xdata', x(:,it), 'ydata', y(:,it), 'cdata', val(:,it));
pause(0.1);
end
Is that the sort of thing you have in mind? If not, perhaps a sketch would help?
