Introduction

This is a set of basic instructions to correlate cores between different holes, written from the perspective of an inexperienced user. This document is intended primarily to handle the case where multiple holes may cover the same interval rather than coring that intentionally covers the same interval, with offset cores as might be done with a paleoceanography leg. This guide also assumes that you are using the Correlator software as an aid in the process.
The goal is to identify the best offsets between holes/cores and if appropriate to define a composite depth scale. There are 3 steps involved in this process:
1. Downloading data from LIMS using Correlator Downloader;
2. Correlation between holes using Correlator based on core measurements to generate an affine table that describes the depth relationships; and
3. Validation and uploading of the affine table to LIMS using SCORS uploader.
After the correlation process is complete, you may be interested in defining splices, but that is outside the scope of this document.

Downloading data from LIMS

Correlator requires CSV format files as the input. Correlator Downloader is a Java program that handles the downloading. There are restrictions on the Correlator input that affect the way you do this:
1. Correlator will only input a single data type once for a given hole;
2. Correlator will only input a single Section file for the entire project; and
3. Correlator is extremely picky about headings in the CSV files.
The current version of the Correlation Downloader program is 12. Earlier versions may not generate files in the correct format. Key things to note here:
1. Select each data type in turn and then click on "Add >>" in the upper part of the window because multiple select does not work at present.
2. Select the samples: this can be done at the level of Hole or Core, i.e. if you select a Hole you get all the data from that Hole, whereas if you select a Core you need to specify each individual Core to download.
3. It is vital that you check "Append core data to hole," otherwise you will get a set of files rather than the single file per data type required by Correlator.
4. The most useful data types are likely to be NGR; Magnetic Susceptibility; a*, b*, L*; and possibly RGB. These are less affected by recovery and fracturing than other properties such as P-wave velocity and density.
5. You might also choose to trim the data acquired at section ends – the downloader allows this independently for top and bottom. If you wait to do this later, Correlator will allow you to trim the top but not the bottom.
6. Finally, click on "Download correlation data".
7. Before going further you need to combine the contents of the Section files into a single CSV file. The file name does not matter but this must include all of the section information at the site; the column headings are critical and Correlator will complain if any of them are different from what it expects.

Using Correlator

These notes are written using Correlator 2.1rc2.
First off, start the program running. This brings up a screen entitled "Data List" and few hints on how to proceed. There are a couple of main menu items, and a floating tool bar.
Correlator stores files at a location controlled by the "Data Repository Path," which can be accessed from the main File menu, so configure this first.
If Correlator appears to freeze, you can often resolve this by using the tool bar to click between Data Manager and Display.
Note that there are several videos illustrating the use of Correlator, and also a User Guide. However these come from an older version and have some features that no longer exist. The videos also tend to go too quickly over important aspects of the process.

Input data

First, click on "Root" to generate a menu and select "Add new data". Select the data file to import, and click OK. This will bring you to a screen called "Generic Data" that shows the first 30 lines of the file parsed so you can check that it is being read correctly.
The first column here is critical, and may be filled or may be empty. If it is empty, you need to specify the data type by left clicking on the heading "Data Type." You will get a set of options that are already defined.
If it is a new data type, select "User define", input the name and probably also check the box "Register," which will add the new data type to the set of options. Now click "Import".
This brings you back to the "Data List" screen, but there should now be an entry for the Expedition, Site, and Hole for which you just imported the data. Do this as many times as you have input files. There is a bug when your data type is Susceptibility that will not return you to the Data List screen automatically. In this case use the tab at the top of the main window to go back to Data List.
Finally, you need to populate the Section Summary. Right click on "Section Summary" in the Data List window and choose "Import Section Summary". If you already have one, you must delete it first before adding new lines. Correlator cannot cope with having more than one or merging a new one. Once you get a window saying that Correlator is happy, click "Import" otherwise fix whatever it is complaining about.

Displaying and shifting data

Go back to the Data List window. You should now be able to see all the data types you are going to use for all the holes that you are going to link.
Note particularly the column headed "Enable;" the status here controls which datasets you see in the main compositing screen. If you right click on the entry next to a dataset you can see the options (Enable/Disable/Delete/Export). Set the data you want to use to "Enable" and those you do not want to use to "Disable".
Now you can highlight an item, right click on it, and select "Load". An item here can be an individual dataset for a single hole, a data type (e.g. NaturalGamma), or the entire Site.
You are now in the main Composite window. Note that you have a set of tabs on the right hand side of the window that allow you to move around: Preferences and Filter, allows you to trim the data (core top only) and apply a range of smoothing options.
To test a tie between features in two holes, shift-left click on the master (which will stay at the same depth), and then shift-left click on the corresponding feature. In the Evaluation window on the right hand side you can see (and control) a cross-correlation function between the two curves. Clicking on the circle at the left hand end of your pick line lets you move the pick by hand. In the Composite panel you can apply the shift when you are happy; there are a number of options for this that are generally self-explanatory. When you are happy overall, you can click "Save affine table" or just reselect the Data Manager in the tool bar.

Extending a hole

This is a slightly painful process if it happens. Essentially you need to delete the existing files for the hole: Select a data item in the Data List screen, then right click on it and select "Delete" (note that what gets deleted appears to be what is selected, not what you right click on which is different behavior from many other options). Now you can import a new file containing the extended data. Delete the existing Section Summary, and import an updated one. And continue as before.

Exporting results

When you are happy with your fit, go back to the Data Manager. Click on "Saved Tables" to see the list of tables completed, select with the left mouse button, and then right click to access the menu for Export. This will export a CSV file formatted according to the requirements for the LIMS upload.

LIMS upload

This is the easy bit – find a programmer/technician and give them the affine table file. They can then attempt to upload it into LIMS and make the composite depth scale available through LORE as an Alternate Depth Scale.