The distributions of precipitations have been changing across the Mediterranean region in the last 70 years… but not always as we expect !

Climate change is known to have consequences on both mean and extreme precipitations, with potentially threatening impacts on societies. The Mediterranean region is particularly affected, being a hotspot of temperature and precipitation changes: this region is expected to get much drier, but with more extreme rainfalls. Though the two extremes of precipitations (absence of rain -drought, dry spells etc. - and very heavy rainfalls) are crucial and well-studied, the evolution of the rest of the rain distribution has been quite overlooked in the literature. Still its study might help to get a broader and more coherent picture of the evolution over the last decades. In the Mediterranean, we commonly expect a “water cycle paradox”, i.e. decreasing mean annual precipitation (“drying”) while very heavy rainfalls intensify.

In this presentation, we look at how the whole wet-days precipitation distribution changes, in the Mediterranean region over the recent past. We use reanalysis data (ERA5) covering the whole 1950-2021 period at daily timescale. We study the trends of the rain percentiles and their statistical significance over the last 70 years.

• We see indeed sub-regions where the “water cycle paradox” is happening, such as the Iberia peninsula as a whole or more specifically Andalusia. For those, the quantile trend curve is in a “U-shape”, with decreasing rain quantiles up to a given threshold (“inversion quantile”) and then an increasing distribution tail. This “inversion quantile” can vary a lot from one place to another.

• However we also find that the “U-shape” trend behavior is not the norm for the Mediterranean region: the situation is more complex. In fact, two additional behaviors are observed according to the location: some regions where the whole rain distribution decreases, and others where it all increases (which is a behavior more expected in Northern Europe or over the oceans). We give a map of the “U-shape” regions and of those two supplementary behaviors, and test its robustness.

• By modeling the rainy days distribution with a simple Weibull law (2-parameters), we manage to get an analytical criterion for the type of rainfall percentiles trend behavior. This Weibull model also enables us, for regions having the “U-shape” behavior, to derive an analytical expression for the “inversion quantile”.