Links between circulation indices and precipitation in the Mediterranean in an ensemble of regional climate models

Abstract

Spatial and temporal variability of precipitation in the Mediterranean is related to atmospheric circulation patterns such as the North Atlantic Oscillation (NAO), the Mediterranean Oscillation (MO) and the Western Mediterranean Oscillation (WeMO). This study examines ability of an ensemble of 12 regional climate model (RCM) simulations to reproduce observed links between these circulation indices and precipitation, as well as how these links may change in the late twenty-first century. We focus on the winter season and differences in precipitation amounts on the highest and lowest 25 % of days according to a given index. The relationships are evaluated against the E-OBS data set for . The observed pattern of differences in precipitation between positive and negative phases is generally similar for MO and NAO, which relates to the high correlation between these indices. Most regional climate models (RCMs) simulate links between the circulation indices and precipitation over most of the Mediterranean area reasonably well, especially for the MO and WeMO indices. The RCM with the largest deficiencies in reproducing the links is HadRM for all indices. The spatial patterns of differences in daily precipitation under positive and negative phases of the circulation indices for the future scenario are similar to those for the control climate for all indices. This suggests that NAO, MO and WeMO are likely to play similar roles in affecting precipitation in the Mediterranean also in the future. However, increased NAO and decreased WeMO index, projected in most examined RCMs for the late twenty-first century in winter, may affect overall precipitation patterns.

Complexity and predictability of the monthly Western Mediterranean Oscillation index

“…Peninsula López-Bustins et al, 2008; Gonzalez-Hidalgo et al, 2009), 85 being also used in other regional climatic applications (Azorín-Molina and Lopez-Bustins, 2008; 86 Sánchez-Lorenzo et al, 2009;Vicente-Serrano et al, 2009;Ouachani et al, 2013; El Kenawy et 87 al., 2013;Beranová and Kyselý, 2015;Ríos-Cornejo et al, 2015). (Figure 1b), strongly suggests that this decreasing tendency could be mainly 107 linked to the behaviour of WeMOi at spring and summer seasons.…”

“…The WeMOi was proposed by Martín‐Vide and López‐Bustins () to detect atmospheric circulation patterns related to rainfall shortage or excess affecting the eastern Iberian Peninsula (López‐Bustins et al ., ; Martín‐Vide et al ., ; Gonzalez‐Hidalgo et al ., ), being also used in other regional climatic applications (Azorin‐Molina and Lopez‐Bustins, ; Sánchez‐Lorenzo et al ., ; Vicente‐Serrano et al ., ; El Kenawy et al ., ; Ouachani et al ., ; Beranová and Kyselý, ; Ríos‐Cornejo et al ., ). The WeMOi is defined as the difference between the normalized monthly barometric series at San Fernando (Spain) (36°17′N, 06°07′W) and the normalized monthly barometric series at Padova (Italy) (45°24′N, 11°24′E), with average and standard deviation being derived from the 1961–1990 period.…”

Evaluation of Historical CMIP5 GCM Simulation Results Based on Detected Atmospheric Teleconnections

“…A growing number of studies focus on winter circulation patterns in the NH in the Euro-Atlantic region [9][10][11][12]. Teleconnections and their effects on atmospheric variables were also studied in the region of the Mediterranean Sea both by using regional climate models (RCMs) [13,14] and GCMs [15]. Recently, the connection between teleconnections detected in the Eurasian region (e.g., the East Atlantic/West Russia and the Scandinavian patterns) and distant areas in Asia were examined [16] while in the region of the Pacific Ocean, among others, the North Pacific Oscillation (NPO) was analyzed [17].…”

Process‐based evaluation of the VALUE perfect predictor experiment of statistical downscaling methods

“…Currently, there is a growing number of studies emphasizing the need for process‐based model evaluations (e.g., Flato et al ., ), which allow for a better understanding of model errors and their cause. Broadly speaking, there are two approaches, the process isolation and the component isolation techniques, also called “regime‐oriented” and the “physical component isolation.” Within the former, model results are averaged within categories that describe physically distinct regimes of the system, namely circulation regimes and thermodynamic states (Flocas et al ., ; Maraun et al ., ; Kjellstro m et al , ; Marteau et al ., ; Addor et al ., ; Beranová and Kyselý, ), instead of being averaged over space or time (e.g., global averages or yearly mean values). The “physical component isolation” technique isolates model components or parameterizations in off‐line simulations that are usually compared against specific observational field campaigns.…”