Basically, we can take a set of numbers and break them into quarters. Just like we do with the months in a year.
Interquartile is the set of numbers between Q1 and Q3, the numbers in the middle of the dataset. So if the models are putting out numbers that are very spread out and uncertain like in your example, the Interquartile Range will be high. If the models are all putting out similar numbers, the difference between Q3 and Q1 will be low, so making a temperature forecast might be easier.