Selecting a Time Period

You should now feel comfortable with selecting data for a specific region and for a specific variable. The last parameter to which data is typically constrained is the time period. You saw in a few of the examples in previous sections that time period selection typically occurs along with other steps. We have kept this step separate here to ensure that all of the methods by which a time period can be chosen are covered. As in the previous section, these techniques are the same for both station and gridded datasets.

Note: When using station data, it is often best to select your desired station(s) before selecting a time period.

Selecting a Continuous Time Period

Start at the Reyn_Smith dataset main page*.

Example: Select data for the time period November 17, 1982- December 29 , 2001.
The weekly SST variable is used in this example.

Method 1: Data Selection Link
Select the "Data Selection" link in the function bar and then the "weekly SST" link. CHECK
Note that the default values in the time grid text boxes are the limits of the dataset.

Enter "17 Nov 1982 to 29 Dec 2001" in the appropriate text box in the Setting Ranges Table.
Select the "Restrict Ranges" button.
CHECK
Note that the limits of the time grid in the Data Selection box match the closest Wednesday to your selected time limits as weekly data are stored on the date of the Wednesday of each week.

Select the "Stop Selecting" button. CHECK
You have now selected weekly SST data for the time period November 17, 1982- December 29, 2001 and can see these selections in the source bar.

Method 2: Expert Mode
In this example, we can do all of our selections in expert mode.

Go back to the Reyn_Smith dataset main page.
Select the "Expert Mode" link in the function bar.
Note how the dataset is represented at the very top of the page.
Enter the following lines in the text box below the text already in the box.

.weekly .sst
T (17 Nov 1982) (29 Dec 2001) RANGE

Click "OK". CHECK
You now have the same selection results as those found with Method 1. Note that your selections are now in the source bar as well as at the top of the page.

Selecting a Discontinuous Time Period

Let us use the GLOBALSOD* dataset for this example.

Example: Create a time series of data on Sept. 3 -Sept. 4 including the years 1995-1999.

Enter expert mode. CHECK
Enter the following line in the text box below the text already there.
T (3-4 Sep 1995-1999) RANGE
Click "OK". CHECK
Note that your date selection is not only seen in the source bar, but in the time grid information at the bottom of the page as well. Let's look at another example of this technique with the Reyn_Smith dataset.

Example: Create a time series of September data including the years 1990-1999.

Go back to GLOBALSOD dataset main page.
Enter expert mode.
CHECK
Enter the following line in the text box below the text already there.
T (Sep 1990-1999) RANGE
Click "OK". CHECK
Again, note that your date selection is not only seen in the source bar, but in the time grid information as well.

Example: Create a time series of data that includes every Monday in 1999.

Go back to GLOBALSOD dataset main page.
Enter expert mode.
CHECK
Enter the following lines in the text box below the text already there.
T (4 Jan 1999) (31 Dec 1999) RANGE
T 7 STEP

Click "OK". CHECK
The STEP command selects every n time steps, where n is the indicated period (e.g., 7), beginning with the first available . Therefore, we specified our range to begin on Jan 4, which is a Monday.

With the techniques highlighted thus far, you should have an excellent understanding of how to navigate through the pages the of the Data Library and how to make basic data selections. The next logical step then would be to manipulate the data statistically or arithmetically, create plots of these data, or download the data you have selected. Part III focuses on these tasks.