You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While the regrid function is useful, it would be great to have a few extra tools for putting data onto a new ocean grid. Specifically, I propose some version of the following:
bin3d: takes sparse ocean data, averages all points which lie within an ocean grid cell, outputs the average and stdev, with missing for all grid cells where no data exist
within this function you could have four keyword arguments:
interph - interpolates horizontally along the grid and inpaints NaNs around the edges, to completely fill the grid horzontally
interpv - same as above, but vertically
interphv - first interpolates horizontally, then interpolates vertically to fill in any 'levels' of the model which weren't filled in by hinterp because no data existed at that level
interpvh - same as above but vertically interpolates first
Incidentally, I've thought a lot about how to plot GEOTRACES data (and same may be true for other ocean datasets). I think that oceanographic data is typically sampled in higher vertical resolution than horizontal resolution, compared to the relative changes between stations vertically and horzontally. Thus, the most 'true' way to plot sections is by interpvh onto a high-resolution grid (or at least high resolution in the vertical, the horizontal resolution doesn't matter much. That makes the sections plot 'truth' at each data point where samples were collected (background field matches actual observations).
Thus, this functionality may have some useful plotting applications, in addition to being useful for binning data for other reasons.
The text was updated successfully, but these errors were encountered:
While the
regrid
function is useful, it would be great to have a few extra tools for putting data onto a new ocean grid. Specifically, I propose some version of the following:bin3d: takes sparse ocean data, averages all points which lie within an ocean grid cell, outputs the average and stdev, with
missing
for all grid cells where no data existwithin this function you could have four keyword arguments:
interph - interpolates horizontally along the grid and inpaints NaNs around the edges, to completely fill the grid horzontally
interpv - same as above, but vertically
interphv - first interpolates horizontally, then interpolates vertically to fill in any 'levels' of the model which weren't filled in by hinterp because no data existed at that level
interpvh - same as above but vertically interpolates first
Incidentally, I've thought a lot about how to plot GEOTRACES data (and same may be true for other ocean datasets). I think that oceanographic data is typically sampled in higher vertical resolution than horizontal resolution, compared to the relative changes between stations vertically and horzontally. Thus, the most 'true' way to plot sections is by
interpvh
onto a high-resolution grid (or at least high resolution in the vertical, the horizontal resolution doesn't matter much. That makes the sections plot 'truth' at each data point where samples were collected (background field matches actual observations).Thus, this functionality may have some useful plotting applications, in addition to being useful for binning data for other reasons.
The text was updated successfully, but these errors were encountered: