further improvements to time-dependence of historical GC-Net elevations

[jasonebox]

we can/should improve the approximations *_info.csv (magenta line in figures) e.g. by

1. obtaining more time series from Abbas Khan
2. reconciling the 3-6 m offsets between the Jacobsen 2-frequency results (cyan circle) and the monthly averaged 1-frequency AWS data *red squares). I had a chat with JJ after my presentation yesterday and he suggested an issue something like atmospheric influence on 1-frequency results.
3. incorporating more data, e.g. PRODEMS from Mai Winstrup that are in the GEUS dataverse