This option allows you to restrict the correlations to a specific season. Many effects vary greatly with the seasonal cycle, and it is rarely a good idea to combine all seasons.

In the Climate Explorer, the season is defined by a starting month and the duration of the season. The starting month is

• all months separately, this gives 12 results, or
• a single calendar month (each January, February, ...), or
• all months together, this gives one result.

Some of these options are not supported in specific contexts.

The length of the season can be interpreted in two different ways:

• an average over the months of the season, or
• all months of the season considered separately

If there are two datasets the lengths of the seasons can be chosen separately. Again, not all combinations make sense.

• To compute a property of all winters in the Netherlands, set the starting month to [December], keep the method on [averaging] and select the length of the season as [3]. This enters a single number for each year, the average over December, January and February.
• To get twelve 3-month averages, change the starting month to [all]
• To study all NH winter months, set the starting month back to [December] and change the method to [selecting]. This enters three numbers for each year into the calculation, the values for January, February and December.
• To compute a property of the anomalies with respect to the seasonal cycle, put the starting month to [together], check the box anomalies and make sure the length of the season is equal to [1].