search

ggebbie / OceanGreensFunctionMethods.jl

Green's Function Methods for Tracer Timescales and Pathways in Ocean Models
MIT License
4 stars 2 forks source link

RTD not normalized by number of boxes #33

Open ggebbie opened 3 weeks ago

ggebbie commented 3 weeks ago

inconsistent with manuscript? inconsistent with MATLAB code?

ggebbie commented 2 weeks ago

Hi @ThomasHaine and @MaceKuailv , Please take a look at this possible inconsistency below.

RTDs as computed in this package, e.g., here https://github.com/ggebbie/OceanGreensFunctionMethods.jl/blob/87707eb3644d899cd48dfb5eda24f301204a2613/notebooks/pedagogical_box_model.jl#L670, match Figure 11 in the manuscript preprint.

In the source code here: https://github.com/ggebbie/OceanGreensFunctionMethods.jl/blob/87707eb3644d899cd48dfb5eda24f301204a2613/src/greens_functions.jl#L115 there is no division by N as described in equation 94 of the manuscript.

In the MATLAB code, function RTD in build_Greensfns also does not have an explicit division by N. Is there an inconsistency here?

MaceKuailv commented 2 weeks ago

Do you have a plot of the RTD calculated with this package (without 1/N)? It does seem like compared to the TTD lines in figure 10, the RTD lines are shifted upward a bit.

ggebbie commented 2 weeks ago

Here are two figures from the Pluto notebook that correspond to the two panels in Figure 10. I didn't detect any offset to the manuscript. I didn't take another look at the MATLAB result.

screenshot-2024-10-04-11:58:56 screenshot-2024-10-04-11:59:18