apply slotgemiddelden to havengetallen and gemiddeldgetij

[veenstrajelmer]

The havengetallen and gemiddeldgetij should be corrected with slotgemiddelden. This corrects these kenmerkendewaarden from being representative to the middle of the 10-year period to the end of the 10-year period. For gemiddeldgetij, this correction happened before via the A0/MIDD in the component file, which is not used anymore since we get gemiddeld getij directly from measurements. This means we also compute gemiddeldgetij from 10 instead of 4 years, but this does not matter too much. For havengetallen, this correction was also mentioned in old literature. Also, if we do not correct both for slotgemiddelde, the gemiddeld getij signal might be skewed. However, there are no slotgemiddelden available for spring/neap hwlw's.

The computation of slotgemiddelden was made less complex in #156. The slotgemiddelde is mainly to convert these KW from representative for the middle of the period to representative for the end of the period.

KD 20-06-2024: Oorspronkelijk werden de waarden van gemiddeld hoogwater en laagwater over de laatste 10 jaar zoals gepubliceerd in Tienjarige Overzichten als standaardwaarden gebruikt. In de jaren 1930 was het tijverschil in de westelijke Waddenzee sterk toegenomen door de afsluiting van de Zuiderzee. Verder was men inmiddels zich bewust van trends (men sprak destijds nog vaak van bodemdaling i.p.v. zeespiegelstijging). Het slotgemiddelde was de aflezing van een trendlijn na wegfiltering van de cyclus. Later: aflezing van het lineaire deel van een gefitte functie die de som was van een rechte en een sinusoide met opgelegde periode en fase aan het eind van het laatste decennium. Teruggaan naar het gemiddelde over het laatste decennium lijkt me daarom geen verbetering.

For gemiddeldgetij, it seems we can just replace A0:

    # overwrite A0 with slotgemiddelde for waterlevel (of entire timeseries)
    slotgem_dict = calc_slotgemiddelden(df_meas=df_meas,
                                        min_coverage=min_coverage,
                                        clip_physical_break=True)
    slotgem_wl = slotgem_dict["wl_model_fit"].iloc[-1]
    comp_av.loc["A0", "A"] = slotgem_wl
    comp_frommeasurements_avg.loc["A0", "A"] = slotgem_wl

Todo:

• pass complete measurement series throughout code #166
• deprecate debug argument for calc_gemiddeldgetij #169
• align code with min_coverage and modular newyear clipping #167
• consider deprecating clip_physical_break keyword for kw.calc_slotgemiddelden() and kw.calc_overschrijding() (but always do it, and log it)
• run calc_slotgemiddelden (including min_coverage=min_coverage and clip_physical_break=True) from within calc_havengetallen and calc_gemiddeldgetij.
• replace A0 in gemiddeldgetij with slotgemiddelde (wl)
• correct havengetallen with slotgemiddelde (HW/LW) >> this is not possible for HW/LW for spring/neap tide, so cannot do this generically.
• update tests
• compare to gele boekje or Gemiddelde Getijkromme 1991.0

[veenstrajelmer]

When not correcting the gemiddelde getijkromme A0 with the slotgemiddelde for waterlevels we get this:

When applying the correction, we get this:

When alternating between the images above we can see that only the "ruwe krommen" change. This makes sense since the eventual krommen are corrected with mean HW/LW for the spring/neap/mean tide (from kw.calc_havengetallen()). So it seems it does not make a lot of sense to only correct the gemiddeldegetijkromme with the slotgemiddelde. In the past, the gemiddelde getijkromme used slotgemiddelde as A0 and now we use the mean of the last 10 years. Since we do scaling with havengetallen, the impact of doing this is very limited.

In principle we could also correct the havengetallen with the slotgemiddelden. However, we would need a slotgemiddelde for hw/lw for spring/neap tide also, this is not available. Alternatively we could correct the mean from the 10-year period with the slotgemiddelde for the entire period. However, we would have to derive this from waterlevel measurement timeseries, while within the kw.calc_havengetallen() we only use waterlevel extremes.

Also, when reading Gemiddelde Getijkromme 1991.0, it seems that the gemiddelde getijkromme was not corrected for the slotgemiddelde anyway (paragraph "Onbeinvloed gebied"), but instead only corrected with mean HW/LW from the havengetallen. This is what is implemented already.

The only change that would make historic sense is to replace A0 with the slotgemiddelde. As described above this would only have very little effect and almost nothing on the scaled getijkrommes. Therefore, it seems wise not to overly complicate things for now and do not correct gemiddelde getijkrommen and havengetallen with slotgemiddelden. This can be reconsidered later.