about OSL dating in OxCal

[Junjie Ren]

Hi Chris,
    I have several radiocarbon ages in a section. But I can't find any radiocarbon sample in a unit. I collected a OSL sample to date the unit. For example, the OSL age is 4.5±0.7 ka. Can I input this OSL age as C_Date (4500,700) in my oxcal model?
Thanks.
Junjie

--

Junjie Ren
Currently as a visiting scholar in Oregon State University
Dept. of Geosciences, 255 Wilkinson Hall, Corvallis, OR 97331-5506
Assistant Professor
Active tectonics & Earthquake

Institute of Crustal Dynamics, China Earthquake Administration
Xi San Qi ,Hai Dian  Beijing 100085,China
86-10-62846730

[Christopher Ramsey]

Junjie

Yes - but only if you have the options set to accept dates in BP - a clearer way of entering this information is:

Date("OSL-ID",N(calBP(4500),700));

You also need to check if the OSL date is in calBP (before 1950) or if it is before the measurement date. If the latter and if this date was 2012 you could use:

Date("OSL-ID",N(2012-4500,700));

Best wishes

Christopher

> --
> You received this message because you are subscribed to the Google Groups "OxCal" group.
> To post to this group, send an email to ox...@googlegroups.com.
> To unsubscribe from this group, send email to oxcal+un...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/oxcal?hl=en-GB.

[Junjie Ren]

Thanks.

[Ana Cathena]

Dear Christopher,

I input my OSL dates by editing the script to match the format you offered below and also tried the C_Date option, without success. You mention in the manual that Oxcal is willing to treat OSL Dates as historical information. Is there a script I can use to input already calibrated OSL dates and not have Oxcal expand the uncertainty?

Thank you for taking the time to answer all the questions here.

Ana

[Christopher Ramsey]

Ana

Can you post a few lines showing how you are entering the dates so that I can see why it is not working. To answer your other question it is fairly easy to get Excel to generate the necessary code for you using the CONCATENATE command - but obviously you have to have a command which is working for your data first.

Christopher



On 14 Jan 2013, at 09:02, Ana Cathena <anaca...@gmail.com>
wrote:

> --
>
>

[Ana Cathena]

Thank you for your rapid response Christopher,

I tried inputting as you directed in this thread:

  Date("OSL-7",N(calBP(900),100));
  Date("OSL-8",N(calBP(4600),400));
  Date("OSL-1",N(calBP(5400),700));

For OSL7, this method output a 95.4% probability of the age being between 700 and 1100 years BP. The plot title came up OSL7 N (1050.5,100).

And I tried using the C_Date input for the same sample which generated this line of script.

  C_Date("OSL 7", 900, 100);

The output is a 95.4% probability of the age being between 851 and 1251 years BP. The plot title is C_Date (900,100).

I have never used the CONCATENATE function in excel. Looks fairly straightforward. Then when I import, I just have to figure out which mode to select from the pull down menu. That seems to be my biggest challenge. Once I get these OSL's input correctly, I am going to try and mimic your input from the 2009 article.

I also tried this C_Date("Historical informaton", 900, 100). And the curve doesn't match the date.

Hmmm.

Ana

[Christopher Ramsey]

Ana

Internally OxCal uses astronomical years as its time base - that is 900 is the start of the year 900. This provides a well-defined timescale where more positive numbers are later in time as you would normally expect for a system of units. It is also sufficiently different from all of the different dating terms that tend to get used (cal BP (meaning before 1950), BP (meaning radiocarbon years, before present - as 1950, or BP as in before measurement) and b2k, not to get confused with them.

Without any special options set, both Date and C_Date assume you are talking in these terms. For your convenience you can use the calBP function to turn calBP dates (meaning before 1950) into astronomical years for you - but for reasons we'll come on to for OSL you may not wish to do this.

On 14 Jan 2013, at 18:48, Ana Cathena <anaca...@gmail.com> wrote:

> Thank you for your rapid response Christopher,
>
> I tried inputting as you directed in this thread:
>
> Date("OSL-7",N(calBP(900),100));
> Date("OSL-8",N(calBP(4600),400));
> Date("OSL-1",N(calBP(5400),700));
>
> For OSL7, this method output a 95.4% probability of the age being between 700 and 1100 years BP. The plot title came up OSL7 N (1050.5,100).
>

Yes - so this is right the 1050.5 is the middle of the year AD1050 which is 900 years before AD1950 - and the range is as you would expect.

> And I tried using the C_Date input for the same sample which generated this line of script.
>
> C_Date("OSL 7", 900, 100);

This is giving the start of AD900 which is 1050.5 years before the middle of AD1950 - hence the range that you see. If you want this to give the same value as the method above you can either use:

C_Date("OSL 7", calBP(900), 100);

or you can set up C_Date to accept BP dates. I don't really recommend this as it means you need to specify the option with the code whenever you distribute or publish it. The whole code would then look like:

Options()
{
BCAD=FALSE;
};
Plot()
{

Date("OSL-7",N(calBP(900),100));
Date("OSL-8",N(calBP(4600),400));
Date("OSL-1",N(calBP(5400),700));

C_Date(900, 100);
C_Date("Historical informaton", 900, 100);
};

Now I assume that if you are using the actual OSL age this is from measurement rather than 1950 so it makes sense not to use calBP at all which really ought to be reserved for calibrated radiocarbon dates (where there is some reason for 1950 to be used as a pivotal year). For OSL I think the most straightforward way to enter this (without setting the options above is either as):

Date("OSL-7",N(2012-900,100));

or

C_Date("OSL-7",2012-900, 100);

where 2012 is the measurement year and 900 is the age at that point in time.

>
> The output is a 95.4% probability of the age being between 851 and 1251 years BP. The plot title is C_Date (900,100).
>
> I have never used the CONCATENATE function in excel. Looks fairly straightforward. Then when I import, I just have to figure out which mode to select from the pull down menu. That seems to be my biggest challenge. Once I get these OSL's input correctly, I am going to try and mimic your input from the 2009 article.

so in Excel you would have something like:

=CONCATENATE("Date(",CHAR(34),A1,CHAR(34),",N(",B1,"-",C1,",",D1,"));")

where your label is in A1, your year of measurement in B1, your age in C1 and your uncertainty in D1. CHAR(34) is the quote character ".

You can then copy the text from excel - use [View > Code] in OxCal and paste this code inbetween the "Plot(){" and "};" to get something like:

Plot()
{
Date("OSL-7",N(2012-900,100));
Date("OSL-8",N(2012-910,100));
Date("OSL-9",N(2012-920,100));
};

Best wishes

Christopher

> --
>
>

[Ana Cathena]

Aha. Of course! :) Thank you so much for taking me through these basics and the program's internal workings. I have been hurried, tired, and unwell trying to push toward a deadline. I got myself all turned around. Knowing I'm highly visual, I might have drawn myself a picture in retrospect.

Best to you too. Thanks also for your patience.

Ana