Tidal Analysis

[felicio93]

In continuation of the discussion we had this morning, I previously used utide for my tidal analysis, which initially seems to be producing reasonable results for my 4 month simulations. e.g.:

K1 for the Bering Sea from utide vs. literature:

Also comparing at coops gauges (Atka):
coops:

utide:

However, I heard from @gseroka and later form @acerrone3 that there are some concerns with utide (and potentially other packages as well).

I am creating this space so we can all learn from the tidal experts on our team, share our findings, and discuss best practices for tidal analysis and related topics.

[felicio93]

@acerrone3, you mentioned you guys have compared different packages. Could you please share some of your findings here?

[gseroka]

@felicio93 @acerrone3 Al can provide perhaps some of the figures/time series that he had produced showing the discrepancies. Below are notes from my meetings with UND last year:
Notes from meetings with UND
Nodal factors

• utide is very off for many constituents
• investigating whether UND screwed up or if there are issues with utide
• use utide in pre-processing to get nodal factors for ADCIRC run (18.61 year cycle)
• utide (Matlab, python) v Zetler (figures) vs pytide (python) vs tidefac (Fortran)
• no confidence in utide- esp important for phase corrections
• contacted utide person, took ttide and added features to it

Summary of diffs between UTIDE, TIDEFAC, TMD, and Pytide... nodal factors, equilibrium arguments. Persisting problems with the M1, M3 and S1.

• currently use nodal factors, equil arguments from UTIDE
• decided to include TMD into ADCIRC- more straightforward to implement than pytide (move away from UTIDE)

From Damrongsak Wirasaet (UND):
It has been a while since I looked at this. If my memory serves me well, the values of nodal factor and equilibrium argument of the S2 and other pure solar from utide are not identical to one and zero. Kowailk states in
in his book (see p. 36 of the doc attached) that `they are identically unity and zero for purely solar terms' (the statement makes sense to me at time when I went through that chapter) . TIDEFAC
and TMD seem to have doubts that the formulas used are correct for M1 and L2. Here is a few lines from TIDEFAC code regarding M1

C** EQUATION 207 NOT PRODUCING CORRECT ANSWER FOR M1
C**SET NODE FACTOR FOR M1 = 0 UNTIL CAN FURTHER RESEARCH
C FNDCST(18)=0.

Note that equation 207 mentioned above is from Schureman (1958).

Best,
Dam
Book2019_tides.pdf