Do ensemble medians exist? Options

[beninbaltimore]

Hi,
I haven't researched this heavily but does anyone know why ensembles are presented as means rather than medians? When I look at the ensemble members on ewall, let's say the SREF, it displays the mean in the bottom right. However, there are 21 members, thus any outliers can have a substantial impact on the mean with a sample size of 21. Medians are less affected by outliers with a small sample size. Are there any medians out there or products available to the public?
Now, I'm not questioning why it's done this way as either I am not knowledgeable enough and/or obviously, the means are shown for a reason. It's just something I've thought about because I do statistical analysis sometimes on small sample sizes, not so much for final results, but more to get a sense of what direction the data is going in. Thanks in advance.

[WeatherMatrix]

Great question -- I haven't seen any operational products with medians, but a Google search shows that it is a concept being (at least) tested in the research community. Looking at "spaghetti plots" showing all the members will give you an idea how far off the outliers are. Maybe because of the way they do the ensembles, the range of all is wide, but none are true outliers?

[beninbaltimore]

Thanks for your response Weathermatrix!