I've recently taken an interest in medicanes. Medicanes are a very interesting phenomenon where tropical-like systems resembling traditional hurricanes can sometimes form in the Mediterranean sea. There are noticeable differences between Mediterranean tropical like systems and normal tropical systems however they resemble tropical cyclones far more than they resemble mid-latitude depressions.

Firstly the requirements for formation are broadly similar, there needs to be a strong thermal gradient between the surface and the upper atmosphere, however medicanes (unlike their tropical counterparts) are able to take better advantage of heat from the air not just from the sea surface; this is probably because at these latitudes:

1) the coriolis force is stronger

2) The tropopause is lower 

3) perhaps most importantly the upper troposphere is much colder

In fact analogies can be drawn between medicanes and periphery Atlantic storms such as TS Epislon, TS grade or hurricane Vince. These systems are arguably far more similar to medicanes than other more southern Atlantic depressions. This difference is most notable in the SST requirement which is more generous for these northern systems.

Approx temperature requirements

Normal Atlantic TC: 26C

Periphery Atlantic TC: 21C

Medicane: 15C

Low wind sheer, like normal atlantic TC is crucially important which is not true of mid-latitude depressions. Medicanes do not have well defined fronts, they have high surface heat flux (much like TCs) and banding more typical of a tropical cyclone including an eye wall and eye. The strongest medicanes will also develop upper level anticyclones, which is a very tropical like feature.

Medicanes do have a much shorter lifespan than normal TCs and even periphery atlantic systems. This is most obvioiusly because the Mediterranean is land locked, and these systems weaken very quickly over land (again like TCs). Also medicanes do sometimes form from extra-tropical lows and are 'upgraded' and develop warm cores (atlantic periphery TCs also often behave in this way) however sometimes they can form entirely from convection not just use convection as a primary 'fuel' source once the circulation is established. 

If we were, for the sake of argument, going to consider the mediteranian as a tropical cyclone basin then it is likely it would be subdivided into two sections:

The west and the east mediteranian. There is a nice analogy that can be drawn between this and the North indian basin (which is divided into the Arabian and Bengal subbasins). The West med basin is by far the most active, there are probably 5x as many tropical-like systems generated in the west than the east; primarily this is because the east (like the arabian) is very dry and less suitable to form tropical-like systems. Note the east basin also includes the black sea which yes, is also capable of forming tropical storms (although it has only happened twice).

In terms of the formation time, the mediteranian hurricane season probably starts in September and ends in January (I am defining the season personally to start on 1st September and end on 31st January). Like the Atlantic, cyclones do occasionally form out of season though it is very rare. It would seem that the early part of the season (particularly september and october) tends to be marginally more active than the later part of the season. 

Since the mediteranian is not officially watched, I cannot guarantee that this review of the 'season' is valid as its only really an interest for me. However I will say that 2014/2015 is probably (and note I don't have an easy frame of reference) a well above average season, with significantly more tropical-like systems than normal. A summary map of the season is given below:

I've tried to stay consistent with the sarrif simpson scale. Blue cyclones 'medistorms' are equivalent to tropical depressions, green 'medicanes' are equivalent to tropical storms, yellow 'major medicanes' are equivalent to hurricanes. Black represents invests or non-tropical lows. I have included G, I think a non-tropical system because it may have exhibited some tropical characteristics, however mostly for means of comparison (since this system has an annoyingly tropical like spiral). G is most likely a cold core warm secluded low, not a tropical-like system. 

In terms of pressure, the values are probably slightly different to what one would expect in the north atlantic and equivalent storms would tend to have slightly lower pressure.

Low end Medicanes (TS) would probably have central pressure of about 990mb compared to 1005mb in the Atlantic

Low end Major medicanes (Cat1) would probably have central pressure of about 975mb compared to 987mb in the Atlantic

Low end Cat2 systems (yes this did arguably happen once in 1996) would probably have a central pressure of about 950mb compared to 970mb inn the atlantic.

Anyway with regards to the 2014/2015 season here are satellite images (mostly IR, some vis) of all the storms (including the imposter G)

So 2014/2015 had 6 'named' storms (+1 annoying non-tropical fully occluded low) including 2 medicanes (tropical storm strength) in probably a well above average season (2015 is very boring in comparison so far!). Incidentally I will do a review in February of the current season. 

Note system B, was actually in the mainstream media because this one was extremely close to being a major medicane. Maximum wind-speeds were probably just a few mph under the 74mph requirement. However there is every possibility that system B was a category 1 equivalent for a short time. The next strongest was system C which was a very low end medicane or very high end medistorm, note the very well defined 'pin hole' eye in the visible satellite image (at this point its centered over the Spanish baleric isles). System D was notable in that it was a truly tiny system, and although it was heavily sheared, the eye wall was very well defined. System E is a marginal system, however unlike G this system probably did have a warm core with circulation and can be considered a weak medistorm. System A had quite high pressure, however that is probably an artifact of there being no buoys nearby, in reality the pressure was probably much lower. System F was also a bit borderline and might be better described as a hybrid. F formed from an unstable extratropical low centered over Italy, some development took place after leaving the shore but it was short lived.

Finally here is a timeline of the tropical cyclones this season. Note the lack of cyclones in September (which actually has the highest probability of cyclone generation), most tropical like systems occurred this year in November and December when the weather conditions were favorable; cold shots of air and frontal systems interacting with the med provided a good environment for tropical like development.

Yes, it is very interesting G. Thanks Q

Thanks everyone for reading. Here is a 2013 summary as-well.

Notably 2013 was a far less active season than 2014 with only 2 'named' cyclones.

To sum up we have something like this.

Year #Medistorms #Medicanes #Major medicanes

2014        4                  2                  0 (1)

2013        1                  1                  0

Brackets indicate borderline cases in 2014 one medicane was virtually cat1 strength with a good chance it actually was at some point. 

When I've done a few more years I might get an indication of what an average season looks like. Again, I suspect 2014 was well above average with 6 'named' storms. 

Incidentally the strongest medicane season on record was probably 1996, including this monster

Probably a catogary 2.

Very good read and detailed post thanks Q 

Excellent work Q, I am impressed 

That 1996 storm is phenomenal what with the width of the CDO and the size of that eye feature. Major Hurricane Danny of the tropical Atlantic hurricane season this year looked pretty similar!

I wonder what the return time is on a storm like that... perhaps ever-shorter if SSTs display an overall rising trend in the next few decades, but then again there may be increased wind shear as well so that could limit the potential.

Its not just that storm in particular, there were 3 well publicized storms in 1996, and given that the only publicized storm of 2014 was the virtual cat 1 system, its likely that there were alot of storms in 1996 that just didn't get publized. I recon something like this is easily possible for 1996:

Medistorms: 6

Medicanes: 3

Major medicanes: 3

Essentially 1996 was the 2005 of the Mediterranean. I wish I could verify this though. Thanks for reading SC,I'll be sure to do a 2015 review in February (although it really is a boring season)