Improve mean yearly wl/HW/LW

[veenstrajelmer]

Old comment about mean wl/HW/LW: "vergelijking yearmean wl/HW/LW met validatiedata Anneke (opgevraagd op 28-04-2022) (nu alleen beschikbaar voor HOEKVHLD en HARVT10, sowieso wl is nodig voor slotgemiddelde), it is clear in the HARVT10 figures that something is off for meanwl, dit gebeurt misschien ook bij andere stations met duplicate times in data_summary_filtered.xlsx (also check on nanvalues that are not nan in validationdata, this points to missing data in DDL)".

Also check KWK_getcheckdata.py for monthly means of ts and ext. Are supplied to PSMSL. If we decide to correct for NAP, we have to communicate it to SLM so they do not duplicately do this. #110

Mean waterlevels were also supplied to PSMSL by RWS, were periods with many missing/invalid values excluded just like with kw.calc_wltidalindicators(df_meas, min_coverage=1)?

[veenstrajelmer]

This will be investigated as part of #110, so this less-properly-documented issue will be closed