Please note that this page has been replaced by a new version with updated maps and more detailed information. We're leaving this old version of the page intact for the moment, so we won't break links to it. But if you come to this page often, please change your bookmark to the new page.
At first glance that's a stupid question. We all know what a mediterranean climate is -- it's dry warm summers and cool wet winters. That's like asking what a mountain is; we all know one when we see one.
Or do we? In the eastern United States, the Appalachians are considered a mountain range. In northern India they'd be foothills. We have a similar problem with mediterranean climates. For example, the specific weather that people call "mediterranean" in much of Europe is quite a bit different from the specifics in, say, California. In fact, the more closely you look at them the more different they start to look.
I'm convinced that many of our struggles to grow bulbs and other plants from various mediterranean climates come from an assumption that everything "mediterranean" needs the same basic conditions.
Lee Poulsen put a lot of work into this subject several years ago, and produced some fantastic charts comparing rainfall in various "mediterranean" cities. When he made the graphs, he found important differences -- many "mediterranean" cities still get significant amounts of rainfall in the summer, and the length of the summer "dry" period varies tremendously. Check out the graphs by clicking here.
To give you an idea of how big the differences can be, compare Genoa, Italy and San Diego, California. Both are generally called mediterranean-climate cities. But Genoa gets 50 inches of rain a year (128 cm) and averages at least 1.4 inches of rain a month in summer (3.6 cm). San Diego gets 11 inches of rain a year (27 cm) and averages 0.1 inches of rain per month in summer (0.25 cm).
If you take a plant native to Genoa and try to grow it in San Diego without irrigation, what do you think will happen? And vice-versa?
So just knowing that something comes from a mediterranean climate is not enough. As plant growers, we need more details on the particular mediterranean climate it comes from.
I decided to see if I could create climate maps of the mediterranean regions, all formatted the same way, so we could easily compare climates from one region to another. Fortunately, there is a huge amount of climate data available on the Internet. Lee kindly pointed me to some of it, and I found a lot more (sources at the end of this note).
There are an almost infinite number of ways you could compare climates, but I decided to focus on two things that kill a lot of mediterranean plants: summer water and winter cold.
When I put those things together, I ended up with 18 mediterranean climate zones, ranging from dry summers with frosty winters (Sacramento, CA) to moist summers with fairly frost-free winters (Adelaide, Australia). Click on the graphic below to see a key to the zones, with some representative cities:
A quick explanation of the terms in the chart:
"Mediterranean desert." The official definition of a desert is anyplace that gets less than 10 inches (25 cm) of rain a year. But some regions that gardeners generally consider to be mediterranean are technically deserts (for example, the southern part of the Central Valley in California). So I created a zone for these areas. Anyplace with 6-10 inches (15-25 cm) of rain a year, almost all of it in winter, qualifies.
"Summer rain." The idea here is to measure roughly how dry the summer is and how long the drought lasts. For the climates labeled "dry," "damp" and "moist," I looked at average rainfall in each of the three driest months of summer. The wettest of these three months is what classified the climate. So, for example, if a city gets rainfall of .1 inch (.25 cm), .1 inch, and .3 inch (.75 cm) in the three driest months, it gets classified as "Damp mediterranean."
Those three zones worked very well for classifying the traditional mediterranean climate areas of California and Chile, but when applied to Europe there were problems. None of mainland France or Italy qualified as mediterranean! So I added two more zones, called Wet 1 and Wet 2. For these, I looked at the average rainfall in the two driest months of summer. Wet 1 has two summer months with less than .9 inches (2.3 cm) of rain a month, and Wet 2 has two summer months with less than two inches of rain (5.1 cm) a month and at least one month in winter that has four times the rainfall of the driest month (in other words, the place is kind of wet in summer, but much drier than it is in winter). This is a more complex classification system than I'd like, but it was the only way to make sure places like Rome and the Croatian coast qualified as "mediterranean."
"Winter cold." I looked at the average low temperature in the coldest month of winter. Places with an average low of 30-40 degrees F (-1 to 4 C) in the coldest month were classified as "cold," places averaging 40-45 degrees (4-7 C) were "cool" and 45-60 degrees (7-15 C) were "mild." This gives a fairly good guide to the likelihood of frost -- places labeled "cold" are likely to get significant frost in most years. Places labeled "cool" are on the border. Places labeled "moderate" should be relatively frost-free in an average year.
Any place that has a winter month with an average low temperature of less than 30 degrees Fahrenheit (-1C) was excluded. You'll get very significant freezes in these places every year. If we didn't have this requirement, huge chunks of central Asia and western North America would be classified as mediterranean, because they have very dry summers. I also excluded any place where the low averages over 60 degrees (15 C); those places are in the tropics (parts of Hawaii are a good example).
Putting it all together. Darkness correlates to temperature: the darker the colors are, the colder a place gets in winter. Hue indicates summer rain: reds and oranges are the driest areas in summer; purples and blues are the wettest. So an area that's dark purple is cold in winter and relatively wet in summer; light pink is a dry, warm desert.
Here is an overall map of the five mediterranean regions, side by side: Comparative climates. There are a lot of blues and greens (meaning mild temperatures with some summer rain) in South Africa, Australia, southern Spain, central Chile (around Constitucion), North Africa, and parts of Greece. There's a lot of dark brown and orange (meaning cold winters with very dry summers) in California, northern Chile, and the Middle East.
Here are the regional maps:
(Note that these images are about half a megabyte each in size. They may be slow to download if you're on a dial-up connection.)
Map of Europe. The interaction of land and water make this easily the most complicated of all the mediterranean regions. One example: Much of the world thinks of North Africa as a desert, but Morocco and Algeria have areas that get more average rainfall than most of California.
Map of Australia. As you'll see, there is a substantial mediterranean-climate area in South Australia. Also note the very thin strip of mediterranean climate along the south-western coast near Esperance. If you go inland even 20 km, the amount of winter rainfall drops off so much that it's hard to classify the local climate as mediterranean.
Map of South Africa. Temperatures and rainfall patterns are somewhat similar to Australia. In both places, there is ocean on the poleward side, rather than land. This appears to result in weather that is a bit wetter in summer and milder in winter.
Map of Chile. There are a couple of Mediterranean-climate areas in Argentina, in the rain shadow of the Andes.
Map of western North America. This started as a map of California, but the Mediterranean zone stretches down into Baja California in Mexico. To the north, there's an enormous area that most people would not call mediterranean, but that actually qualifies if you judge from summer rainfall. Overall, this is probably the coldest of the mediterranean regions, and in California one of the driest.
How the maps were created. In each region, I plotted rainfall and temperature data from up to several hundred weather stations. I then drew lines surrounding all of the stations that fell in the same zones. So all of these maps are approximations; your weather will vary depending on local conditions.
--South Africa and Western/Southern Australia have the mildest climates. Compared to the other mediterranean regions, they are a bit warmer in winter and have less severe droughts in summer.
--California's mediterranean climate may be the hardest on plants. It generally has colder winters and drier summers than the other regions.
--Central Chile's pattern is similar to California's, although a bit milder in many areas.
--Coastal Oregon and Washington have weather that resembles a mediterranean pattern in many ways. Officially, climatologists do not classify them as mediterranean, but for plant-growing purposes I think of them as semi-mediterranean. The same thing applies to south-central Chile.
--Europe is a mix of all the other regions. Spain, southern France, Italy, and Morocco/Algeria all have comparatively moist summers. The Greek islands and the Middle East have very dry summers.
--The mediterranean climate zones don't end where you expect them to. This is a result of the problem I mentioned at the top of this note, the differing definitions of "mediterranean" climate. To put it bluntly, if Genoa is mediterranean, it's very hard to argue that Seattle isn't. Their rainfall patterns and minimum winter temperatures are almost identical (Genoa is actually wetter than Seattle).
I spent a lot of time trying to figure out some clever statistical way to remove cities like Seattle from the map. But you have to do very arbitrary things like excluding any city that has more than 35 inches of rain a year, or that has cool summers. Even with those rules, you end up knocking out some cities that are considered mediterranean. For example, if you exclude Seattle using its cool summers, you also exclude San Francisco. If you exclude Seattle using total rainfall, you also exclude Genoa. In the end, I just let the numbers draw the zones. So these maps cover both mediterranean and semi-mediterranean climates.
What's missing from the maps? The hardest part of making these maps was deciding what information to leave out of them. There are several things I would have liked to include:
Any map complex enough to show all of these variables would be unreadable. The ideal solution would be an interactive online map in which you could add or remove variables by clicking on them. If someone wants to make that, I'd be glad to help. But in the meantime, these maps give us more information than we had before.
I think these maps can help you understand some of the problems we have raising mediterranean-climate plants. For example, here's what I think they're telling me about growing mediterranean plants and bulbs in California:
--Why do so many of my South African Amaryllid bulbs need supplemental water in the summer? Because they are used to getting some summer rainfall.
--Why do many European bulbs naturalize well in gardens that get year-round water, when that kills bulbs from places like California? Because many areas in Mediterranean Europe get more summer rain than any other mediterranean region.
--Why do I have so much trouble growing Australian plants? Because they are used to warmer winters and damper summers than I have in my part of California. Even if I protect them from frost, planting them out in the garden with no supplemental water in summer is likely to put them under enormous stress. No wonder they grow better along the milder, damper coast of California.
--Why have so many South African bulbs become weeds in Australia, but not in California? I wonder if one reason is because the climates in South Africa and Western Australia are much more similar than either is to California's.
Western Regional Climate Center (Click on the map.)
El Dorado Weather
South Africa Explorer (A nice set of clickable maps.)
Maps and commentary by Michael Mace, San Jose, CA