Issue with chron controls and age model relevancy.

[SimonGoring]

Reported on the Neotoma slack channel

Let’s say the user using R is interested in one specific site:

1. When using the get_chroncontrol() for one site, it states in the R package documentation that "the most recent chronology is provided here". The most recent in what sense? When it was created? Here an example of a pollen record that has a radiocarbon default age model and a calibrated default age model:

chenevo <- get_download(get_dataset(get_site("Laguna Chenevo"), datasettype = "pollen"))
get_chroncontrol(chenevo, chronology = 1, verbose = TRUE, add = FALSE)

This returns the radiocarbon age model that was made before the calibrated age model. So the “most recent chronology” is somewhat confusing here.
2) It states that by using [chronology], the chronology can be chosen if there are more than one. So here the user would call chronology = 2.

get_chroncontrol(chenevo, chronology = 2, verbose = TRUE, add = FALSE)

This returns the same radiocarbon age model, not the second calibrated age model! (this was picked by Alistair Seddon when working on our HOPE workflow and not getting the age models we expected).

3. When the user wants to get the best age models for all pollen records for a region, it is basically pulling only the default radiocarbon age models. In the case of Europe it would fail to use the newer updated age models with estimated top ages (Giesecke et al. 2014).

4. If different pollen records have 2,3,4 age models, then the user would have to write out which chronology to use for each site? For a synthesis work like HOPE, and probably other projects as well, this is not doable.

5. Also, the reasoning that each age type will be assigned as a default chronology, is information not being explicitly mentioned in the Tilia manual nor in the R package documentation. This important detail is pretty much in the dark.