[MITgcm-devel] evp

[Martin Losch]

Hi there,

I plan to modify EVP a little in order to make it converge (haha). It turns out that any small change (pulling out a factor etc) changes the results so that only 2 digits of similarity remain in lab_sea.hb87 (the only experiment that uses EVP).

Does anyone object to changing evp-related results? Is there anything important (like ECCO X) relying on EVP? (I hope not)

Martin