Dear Ferreters,
G'day
I am facing a problem and I am not sure about the result that I got is right or not. I computed the ocean heat content trend on neutral density axis and now I want to map it back to the standard z-axis (meter/db). I performed zaxreplace function to do that but the result that I got is not showing any value below 1750 m. However, my all data is till 2000 m.
I suggest plotting (looking at) zedp (it's the depths of isopycnals, right?) at large density levels.
Because the maximum depth of density data points is 2000 m, the maximum depth value found in zedp is smaller than 2000 m. Suppose that density = 28.05 at 2000 m. Where is the depth of the density surface 28.0 ? It's somewhat above 2000 m. Where then is the depth of the density surface 28.1 ? You can't determine that because there is no density value larger than 28.05 there. Ferret refrains from extrapolation. In this way, the depths of all density surfaces are either undefined or smaller than 2000 m.
Therefore, when you map a variable defined on the density coordinates back to z coordinates, zaxreplace finds no values for z = 2000 m.
Because in this backward mapping, you don't have to use the same z coordinates as the original model grid, what about using a finer vertical grid such as . . . 1970 m, 1980 m, 1990 m, 2000 m. Then, even though you still miss data at 2000 m, you'll get values at 1990 m.
Hope this helps,
Ryo