age model with outliers: charcoal and charcoal plus

[Bente Philippsen]

Hi!

I am having trouble with an age model of a site where I both have charcoal samples from fill layers and charcoal that had been embedded in mortar. For the charcoal from fills, I use the Charcoal_Plus model, as the samples can be too old as well as too young intrusions. The charcoal from mortar cannot be younger, but only older than the context, so I use the Charcoal model (I have examples for this at the bottom of this post). It is a site with a lot of activity throughout the centuries, so old redeposited charcoal is very common.

My problem is that one of the charcoals from mortar is treated by the model as being a young intrusion, so the modelled age is much older than the unmodelled age ("Charcoal from mortar 1" in the figure) - with very poor agreement. However, the charcoal cannot be younger than the mortar it was embedded into. How can I implement this in the model, as just using the Charcoal outlier model obviously does not help?

All the best,

Bente

  Outlier_Model("IA",Prior("Charcoal_Plus"),U(0,3),"t");
  Outlier_Model("Charcoal",Exp(1,-10,0),U(0,3),"t");
 

    ...

     R_Date("Charcoal 2", 1838, 32)

     {
      Outlier("IA",1);
     };
     R_Date("Charcoal from mortar 1", 1771, 27)
     {
      Outlier("Charcoal",1);
     };