Help with P_Sequence model for mixed post-/pre-bomb ¹⁴C dates and inversion issues

[nithin kumar]

Hi all,

I’m a beginner using OxCal to model two sediment cores and I could really use some guidance. I’m facing two main issues:

1. Some of the ¹⁴C dates are negative (post-bomb), and others are positive (pre-bomb), which I understand relate to the calibration curve used — but I’m unsure how to handle both together in one sequence.

2. One of my cores seems to have a possible inversion (younger dates appearing below older ones), which makes modeling harder.

Below are the codes I’m using for Core A (4 post-bomb + 1 pre-bomb date) and Core B (3 pre-bomb + 2 post-bomb, possibly inverted).

Core A
Plot()
{
  Outlier_Model("General", T(5), U(0,4), "t");
  P_Sequence("NBA_Core", k=1, tau=0.7, U(-2,2))
  {
    Boundary("Start");
    R_Date("NBA1", -64, 31)     { Curve("Bomb21NH2"); Outlier(0.05); z=1.1; };
    R_Date("NBA7", -782, 26)    { Curve("Bomb21NH2"); Outlier(0.05); z=14.1; };
    R_Date("NBA13", -688, 31)   { Curve("Bomb21NH2"); Outlier(0.05); z=27.2; };
    R_Date("NBA19", -383, 27)   { Curve("Bomb21NH2"); Outlier(0.05); z=40.2; };
    R_Date("NBA23", 171, 2)     { Outlier(0.05); z=48.9; };
    Boundary("End");
  };
};

Core B 
Plot()
{
  Outlier_Model("General", T(5), U(0,4), "t");
  P_Sequence("NB", k=1, g=0.7, U(-2,2))
  {
    Boundary("Start");
    R_Date("NBB23", 421, 32)      { Outlier(0.05); z=88.9; };
    R_Date("NBB18", -547, 28)     { Curve("Bomb21NH2"); Outlier(0.05); z=78; };
    R_Date("NBB12", -987, 29)     { Curve("Bomb21NH2"); Outlier(0.05); z=65; };
    R_Date("NBB7", 569, 28)       { Outlier(0.05); z=54.1; };
    R_Date("NBB1", 292, 28)       { Outlier(0.05); z=41.1; };
    Boundary("End");
  };
};

My questions:

• Is it okay to mix pre- and post-bomb dates like this in a single P_Sequence, using different calibration curves per sample?

• How should I deal with potential reversals? Should I try D_Sequence instead, or split the bomb and pre-bomb phases with a Phase() and Boundary()?

• Are the k, tau, and U(-2,2) values appropriate for sediment cores from southern India (~3–6 mm/year accumulation)?

• Is there a better approach to outlier modeling here?

Any advice would be greatly appreciated!

Thanks in advance,
Nithin Kumar