How to calculate Marine Calibration?

[Jo Morris]

I'm new to using OxCal so sorry if this seems rudimentary, but I have been looking for some easy-to-follow step-by-step instructions on how to use the program and have found no clear instructions so I have been fiddling through on my own. I am now trying to calibrate some shell dates and am getting nowhere, so I am hoping someone can point me towards a good set of instructions or provide me with some instructions.

See my various attempts below where I am trying to replicate these published dates:

                 UnCal Age      CalBP (OxCal 4.2, SHCal 13, Marine 13)

Site 1    34,650 ± 630    36,970–40,270

Site 2   24,420 ± 250     27,680–28,600

I have used the web interface to calculate terrestrial calibrations before but when I try to use the Marine 13 curve the output shows that it is still calculated with IntCal 13 and the results are the same as before I try to use the Marine curve so I'm pretty sure it's not working this way. 

Using the web interface Marine 13 is not shown in the main Curve dropdown.  Under Options and the Curve tab it lists 14C Calibration Curve and Marine/mixed Curve so I assume this is where I set the marine curve to Marine 13.  I have tried to run the calibration but the results still seem to be calculated with the IntCal 13 calibration rather than the Marine 13 calibration and I just can not figure out how to get it to use the Marine curve.

I have also tried entering it via the multi-plots function, but here I am even more lost.  I was able to follow the limited instructions on the OxCal online manual to add R_Date to the Plot (File>New; Insert)

Plot()

R_Date("Site 1", 34650, 630)
R_Date("Site 2", 24420, 250)

With this input I get the below as an output:

                           From       To         %

R_Date Site 1   41203   38133   95.4
R_Date Site 2   29152   27972   95.4

But when it comes to setting a Marine curve there seems to be a number of options and I'm not sure if I am making the right choices.

The first one I tried was to select Curve from the Insert dropdown menu. But when I do this it just has an option for Name and Filename and no clear way to select any of the preset curves.

So the next option I tried was Tools>Options there I can select Marine 13 from the dropdown menu under Calibration Curve when I click the >> button it adds this to the code:

Options()

Curve="marine13.14c" 

Plot()

R_Date("Site 1", 34650, 630)
R_Date("Site 2", 24420, 250)

From this code I get these results:

                             From     To        %

R_Date Site 1   40139   37196  95.4
R_Date Site 2   28346   27240  95.4

Ok so the next option is Tools> Curves tab. There I set 14C Calibration Curve Atmospheric to SHCal 13 and Marine/Mixed Curve to Marine 13 and click the >>

Curve("SHCal13","shcal13.14c")

Plot()

R_Date("Site 1", 34650, 630)
R_Date("Site 2", 24420, 250)

I then get this result.  From the code above this seems like the results are for a SHCal13 rather than for the Marine 20 curve

                           From        To        %
R_Date Site 1   41203   38133   95.4
R_Date Site 2   29152   27972   95.4

The final option I have tried is a combination of these last two options.  Tools>Curves Marine/mixed Curve = Marine 13  14 C Calibration Curve = SHCal 13. Then move over to Tools> Options Calibration Curve = Marine 13.

Options()

Curve="marine13.14c"

R_Date("Site 1", 34650, 630)
R_Date("Site 2", 24420, 250)

Which produces this output:

                             From     To        %
R_Date Site 1    40270  36967  95.4
R_Date Site 2   28605   27675   95.4

These results are the closest to the above-published results but they still don't match.  I do not know if this is because they were calculated in OxCal 4.2 whereas these were done in 4.4, even though the calibration curves are the same. 

The other option I have experimented with is adding a Delta R value but no matter what I add the results do not change.  I know I should use a local Delta R offset, I calculate an average Delta R + error for the region here <http://calib.org/marine>, but I am unsure what to enter for Atmospheric % and Marine %.  When I enter these results into the field even just the Delta R values and leave the Atmospheric set to 100% and Marine % 0+-0.  The results do not differ from the above results.

If I choose Reservior from the Insert dropdown menu I can only insert it if I have the Plot() selected not if I have the Curve selected.  I get this code:

Options()

Curve="marine13.14c"

Plot()

Reservoir(-96, 37)
R_Date("Site 1", 34650, 630)
R_Date("Site 2", 24420, 250)

And these results

Warning! Inappropriate nesting - Reservoir into Plot
                                From       To        %
 R_Date Site 1      40270   36967   95.4
 R_Date Site 2      28605   27675    95.4

These are the same as the other Marine 13 results.  The reservoir has made no effect.

So as you can see I've been trying to work this out but I can not get straight results.  I have been unable to find any clear instructions and I have noone to ask for help.  There are clearly things I am missing when it comes to calculating this so I am hoping someone can offer me some help with this.

[Christopher Ramsey]

In the [Options > Curve] for single dates you need to set Atmospheric % to zero for purely marine use. That will set the Marine % to 100.

The same is true fo [Tools > Curve] when putting the commands into a project.

You will also need a Delta_R value so say that is 100+/-50 you should get:

Plot()
{
Curve("Marine13","marine13.14c");
Delta_R("LocalMarine",100,50);

R_Date("Site 1", 34650, 630);

R_Date("Site 2", 24420, 250);
};

Best wishes

Christopher

> --
> 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 on the web visit https://groups.google.com/d/msgid/oxcal/f40165d4-bcc6-4cf1-bbd4-951593c63ca2n%40googlegroups.com.

[Jo Morris]

Thanks so much Christopher, this is great, it worked!  Can you point me towards some further information about setting the Atmosphere %? For example, does this value change if the shell is from an Intertidal zone or if I am using a local freshwater reservoir offset?

J

[Christopher Ramsey]

For marine samples atmosphere % should be zero - this setting is really intended for organisms (usually humans) with a mixed diet which you would determine in other ways - particularly using stable isotopes.

Freshwater systems don't really have a curve of their own so we have to use the atmospheric curve with an Delta_R value - which again needs estimating in some other way - or left very broad.

Christopher

> To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/0c202e94-d32a-4040-9951-7a84b2520246n%40googlegroups.com.

[Samantha Walker]

Dear Christopher,

I am running into issues with the newest OxCal interface using the Marine 20 curve and my delta offset -- the results are not calibrated properly, whether I do individual dates or a batch (when compared to other programs and just in general). I can't find the atmosphere option in this version and I'm wondering if that is the issue? I've played around with every option I can find...I feel like I'm missing something obvious but I would appreciate any insight!

Many thanks,

Samantha 

[Christopher Ramsey]

Samantha

The options for Marine have been reorganised because the mixed options were often found to be confusing by people wishing to do a pure marine calibration.

To do a marine/mixed calibration you need to do [File > New] this is because for any of these calculations you need the Curve, the Delta_R and the radiocarbon date and it is good to see what you are doing - and check that you are using the right Detla_R!

To set the curves up use the [Tools > Curves] option (this is as before)
Now though you select whether you want an atmospheric, bomb, marine curve or a mixture from the menu.
If you then select "Marine" you will be asked for the Delta_R value
Then you need to enter your radiocarbon dates so press [Insert] and select [R_Date]

Enter all the radiocarbon dates you want to calibrate and then press [Run].

The actual calculations and plots have not changed.

Best wishes

Christopher

> To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/e75ee73a-8c95-40ad-8506-88d40522c76en%40googlegroups.com.

[Samantha Walker]

Thank you Christopher. I repeated these steps as you described but with the same erroneous result. For example, a date of  4215 +/-  20 on a walrus sample using a delta offset of 160 +/- 50 is resulting in a calibrated date of 4850-4650 calBP (95.4%). In Calib, the same date results in 3690-4109 calBP....it shows that Marine20 and the delta offset have are inputted in the log under the R_Dates, but it's as if another curve has been applied instead. 

Many thanks,

Samantha 

[Christopher Ramsey]

That is odd - the input file should look like:

 // Delta_R values updated for Marine20
 Plot()
 {
  Curve("Marine20","marine20.14c");
  Delta_R(160,50);
  R_Date(4215,20);
 };

and the output table should look like the attached.  Without the MCMC (not needed here for a single measurement) the result is 4109-3960 as you are getting from Calib.

Perhaps clear the cache in your browser incase something is not working fully.  I have just updated the code view but this should not alter the inputs or what you are getting on the outputs.

Christopher

To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/7a8a5579-c39d-4815-9f9d-10b123d08d74n%40googlegroups.com.

[Christopher Ramsey]

The most likely explanation is that the R_Date function is being put before the Curve function and so the default IntCal20 curve is being used. This should be clear from the plot which will refer to the default Atmospheric curve. It should also be clear from the table and the input code.

Best wishes

Christopher

> On 16 Nov 2022, at 18:27, Christopher Ramsey <christoph...@arch.ox.ac.uk> wrote:
>
> That is odd - the input file should look like:
>
> // Delta_R values updated for Marine20
> Plot()
> {
> Curve("Marine20","marine20.14c");
> Delta_R(160,50);
> R_Date(4215,20);
> };
>
> and the output table should look like the attached. Without the MCMC (not needed here for a single measurement) the result is 4109-3960 as you are getting from Calib.
>
> Perhaps clear the cache in your browser incase something is not working fully. I have just updated the code view but this should not alter the inputs or what you are getting on the outputs.
>
> Christopher
>
>

> <Screenshot 2022-11-16 at 18.03.26.png>

> To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/ABAA10CA-49B5-4FC5-8719-3902DB67F287%40arch.ox.ac.uk.

[Samantha Walker]

Hi Christopher,

Yes! That was the issue ( R_Date function before the curve function). I have to click the circle under the curve function before inserting the R-Date, which I wasn't doing before. it's working now! 

Thanks for all your help!

Best,

Samantha 

[Christopher Ramsey]

Great - thanks for the feedback. I think you were doing this because when the comment about the updated Delta_R values for Marine20 is added at the start of the file the cursor moves back to the start. That was also true with the previous version. I will see if I can make the cursor return to the end point which is where you would expect it to me normally.

Best wishes

Christopher

> To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/8d9e08dc-a2aa-4896-b190-8f0df516171bn%40googlegroups.com.

[Christopher Ramsey]

I have updated the program now to it asks you to check the Delta_R when an new marine curve is added and then adds a comment before the curve at that point. This should ensure that the cursor no longer goes back to the start of the code which is why you had this problem.

On the assumption that Delta_R might in future be different for each curve this code no longer only asks this for Marine 20 but for any marine curve where a Delta_R is used.

Best wishes

Christopher

> To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/19AAB1FA-FFF4-49CC-8E96-7C2E89DBAFF2%40arch.ox.ac.uk.

[Samantha Walker]

Great, many thanks, Christopher!

To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/5A8D61D1-8F34-48F9-A6E7-B45747E8A86C%40arch.ox.ac.uk.