New question on wiggle match dating - models seemingly not reflecting gap

[Morley Eldridge]

So I have a simple model for three dates on a standing dead tree. The results of a run, however, seem to ignore the known gap between the final two dates.

 Options()
 {
  Resolution=1;
  Curve="bomb21nh1.14c";
 };
 Plot()
 {
  Tree Ring D_Sequence("R17")
  {
   R_Date("R17Inner",190,30);
   Gap(27);
   R_Date("R17Middle",240,30);
   Gap(81,1);
   R_Date("R17Outer",100,30);
  };
 };

All three dates considered individually have possible dates as late as the mid 1950s, but this is logically impossible for all to be correct as the tree grew for at least 116 years. 

The  'R17middle' date has modelled to the late 18th century and the 'R17 Inner' is modelled to this or before, all logical given the gaps.  But I don't understand the modelled "R17 Outer" date having a'post'  probability distribution that includes virtually the entire nineteenth and the first half of the twentieth century. If the 'middle' date ends at about 1800, and there is a known gap of 81 years to the 'outer' date, I would have thought the probabilities for intercepting the curve after about 1880 would be reduced to 0?  What am I not understanding about how wiggle match dating works? I would have thought the gaps were fixed, in which case the 'Inner' and "Middle' dates almost have to be in the mid to later 18th century and the 'Outer' would have to be in the 19th century no later that about 1880.  Is it reasonable to conclude that 1880 is a TPQ for the death of the tree (caused by the harvest of 100% of the tree's bark from one section)?     

[Christopher Ramsey]

I'm not sure what the "Tree Ring" text is doing in the code here - it prevents the model from running. The following:

Options()
{
Resolution=1;
Curve="bomb21nh1.14c";
};
Plot()
{

D_Sequence("R17")
{
R_Date("R17Inner",190,30);
Gap(27);
R_Date("R17Middle",240,30);
Gap(81,1);
R_Date("R17Outer",100,30);
};
};

Works but warns you that DSequence does not allow for uncertainty in gaps - so you really just want:

Options()
{
Resolution=1;
Curve="bomb21nh1.14c";
};
Plot()
{

D_Sequence("R17")
{
R_Date("R17Inner",190,30);
Gap(27);
R_Date("R17Middle",240,30);

Gap(81);
R_Date("R17Outer",100,30);
};
};

If you want uncertainties included then you need to use V_Sequence - but it will only work if the uncertainties are an all the gaps and are significant.

Best wishes

Christopher

> On 10 Oct 2025, at 23:42, Morley Eldridge <morley....@gmail.com> wrote:
>
> So I have a simple model for three dates on a standing dead tree. The results of a run, however, seem to ignore the known gap between the final two dates.
>
> Options()
> {
> Resolution=1;
> Curve="bomb21nh1.14c";
> };
> Plot()
> {
> Tree Ring D_Sequence("R17")
> {
> R_Date("R17Inner",190,30);
> Gap(27);
> R_Date("R17Middle",240,30);
> Gap(81,1);
> R_Date("R17Outer",100,30);
> };
> };
>

> <R17 model run result.png>
> All three dates considered individually have possible dates as late as the mid 1950s, but this is logically impossible for all to be correct as the tree grew for at least 116 years.
> The 'R17middle' date has modelled to the late 18th century and the 'R17 Inner' is modelled to this or before, all logical given the gaps. But I don't understand the modelled "R17 Outer" date having a'post' probability distribution that includes virtually the entire nineteenth and the first half of the twentieth century. If the 'middle' date ends at about 1800, and there is a known gap of 81 years to the 'outer' date, I would have thought the probabilities for intercepting the curve after about 1880 would be reduced to 0? What am I not understanding about how wiggle match dating works? I would have thought the gaps were fixed, in which case the 'Inner' and "Middle' dates almost have to be in the mid to later 18th century and the 'Outer' would have to be in the 19th century no later that about 1880. Is it reasonable to conclude that 1880 is a TPQ for the death of the tree (caused by the harvest of 100% of the tree's bark from one section)?
>

> --
> You received this message because you are subscribed to the Google Groups "OxCal" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to oxcal+un...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/oxcal/8d09b0dd-465f-407e-8256-4b10ac4706ecn%40googlegroups.com.
> <R17plot.png><R17 model run result.png>

[Morley Eldridge]

Christopher

Thank you so much! I had misinterpreted the help webpage 'tree ring sequence' under help with input/models, during my most recent re-reading. "Sequence" is listed separately from "Tree Ring Sequence" there. I'd previously tried several different ways including a simple 'sequence' and D-sequence and V-sequence and trying both 'interval' and 'gap' line alternatives to code. I'd also been bouncing between the input help and the commands pages. Without getting reasonable results. Your correction looks so simple; it is frustrating not to have seen it myself. However, I am gratified that your Oxcal model estimates the Outer date as 1862-1878 and my looking at the plotted dates and considering gaps had come up with 1860-1880 as the most likely, though of course without any probability other than  'substantially less than 95%'. I presume that was the original wiggle match method.

Morley

--
You received this message because you are subscribed to a topic in the Google Groups "OxCal" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/oxcal/dehNvEknK6Q/unsubscribe.
To unsubscribe from this group and all its topics, send an email to oxcal+un...@googlegroups.com.