These usually represent a threshold, such as a large amount of rainfall, which if occurs will lead to undesirable outcome(s) e.g. force cancellation or rescheduling of an event. The Claros model and ALR platform have been developed to deliver predictive probabilities for these thresholds.

There are up to two parameters that can be used. These can be set up to give two separate probabilities or used to create a probability of ‘in between’ or ‘not in between’. Of course, the thresholds are flexible, so they can be used for positive outcomes if more useful.