Dynamic and Thermodynamic Drivers of Precipitation Change in Mediterranean-type Climates

All Mediterranean-type climate regions have experienced recent wintertime precipitation declines, contributing to severe droughts in many cases. Understanding whether these declines are driven primarily by changes in large-scale circulation, atmospheric moisture, or submonthly weather systems is critical for interpreting past trends and anticipating future hydroclimate risk. We use constructed circulation analogues together with a Reynolds-decomposition moisture budget to diagnose the respective roles of dynamic circulation change, thermodynamic humidity change, and submonthly eddy activity in driving these wintertime precipitation trends.

We apply both approaches to observations and reanalyses, multiple large climate model ensembles, and a preindustrial control simulation to understand how these processes regulate moisture convergence and precipitation variability across Mediterranean-type climate regions. Circulation analogue results indicate that observed wintertime precipitation declines are predominantly dynamically driven. However, the thermodynamic drying inferred from the analogue method is stronger than that simulated by large ensembles in all Mediterranean-type regions. Moisture budget diagnostics additionally highlight a substantial contribution from submonthly eddy trends in some locations.

By directly comparing the two frameworks, we highlight that estimates of dynamic and thermodynamic trends can depend strongly on the diagnostic method used. In particular, dynamically driven moisture anomalies and changes in submonthly variability can contaminate thermodynamic estimates derived from both approaches. Using the large ensembles, we show that thermodynamic trends inferred from the two methods can even differ in sign. These results underscore the importance of combining multiple diagnostic methods to more robustly quantify the influence of large-scale circulation and humidity changes on regional precipitation decline.