The majority of the Middle East has an arid climate, making the region susceptible to water scarcity. Annual precipitation can reach up to 500 mm along the eastern Mediterranean coast and exceeds 1000 mm in parts of the Zagros Mountains and the Iranian Plateau but falls below 200 mm across most of the region, reaching as low as 20 mm in its hyper-arid deserts (Fig. 1b, c). Moreover, precipitation in the Middle East experiences strong interannual variability, making the region prone to both flooding and drought, which significantly impact society, the economy, agriculture, and ecosystems (Lelieveld et al., 2012; Barlow et al., 2016; Zittis et al., 2022; Coleman et al., 2026; Sohrabi, 2025). These hydrometeorological hazards can exacerbate geopolitical and social tensions in the region, while pre-existing tensions can make the population more vulnerable to such hazards. For example, the multi-year drought from 2007 to 2011 is believed to have acted as a catalyst for the Syrian uprising and civil war that began in 2011, as this drought caused widespread agricultural failure and displacement of over a million people in the region (Gleick, 2014; Kelley et al., 2015). The Libya flooding of September 2023, leaving more than 13 000 people dead or missing, was caused by heavy rainfall and the breaching of two dams, which were poorly maintained due to armed conflict and political dispute in the Derna region (Armon et al., 2025). Given the large impacts of precipitation deficits and surpluses on society and the environment, it is important to better understand the interannual variability of precipitation in the Middle East and the underlying drivers.

From a global perspective, the Middle East stands out as a region receiving most of its annual precipitation during the extended cool season. More than 95 % of the annual rainfall occurs between October and May across a region stretching from Libya in the west to Pakistan in the east, and from the Zagros Mountains in the north to the central parts of the Arabian Peninsula in the south (Fig. 1a). During the hot summer months of June to September, the region remains almost entirely dry due to the influence of the South Asian monsoon, which induces strong subsidence over the eastern Mediterranean and Middle East (Rodwell and Hoskins, 1996, 2001; Tyrlis et al., 2013). During the transition seasons and in winter, the westerlies and subtropical jet migrate southward and can steer midlatitude weather systems into the region, contributing to most of the precipitation during these seasons. The midlatitude forcing typically manifests as extratropical Rossby waves propagating and breaking toward lower latitudes, whereby the associated upper-tropospheric cyclonic circulation anomalies interact with the tropical low-level circulation (Kahana et al., 2002; Kumar et al., 2015; de Vries et al., 2016, 2018). These tropical-extratropical interactions can direct warm, moist tropical air masses into the Middle East, supporting the development of convective storms and heavy precipitation.

Interannual variability of cool-season precipitation in the Middle East is strongly linked to the El Niño–Southern Oscillation (ENSO). The ENSO warm phase, El Niño, generally leads to an increase in precipitation over the Middle East, while its cold phase, La Niña, leads to a decrease. Previous studies have demonstrated this connection based on various types of observations, such as precipitation and river streamflow data in Iran, Israel, and Saudi Arabia (Kiladis and Diaz, 1989; Price et al., 1998; Nazemosadat and Cordery, 2000; Nazemosadat and Ghasemi, 2004; Athar, 2015; Saghafian et al., 2017), gridded precipitation datasets (Mariotti, 2007; Hoell et al., 2015a; Alizadeh and Mousavizadeh, 2025), climate model simulations (Kumar et al., 2016; Hoell et al., 2018, 2024), and seasonal predictions (Abid et al., 2016). The relationship between ENSO and precipitation in the Middle East is not constant over time but varies on interdecadal scales and within the cool season. While El Niño has been associated with a consistent precipitation increase throughout the 20th century, the link between La Niña and reduced precipitation becomes apparent only from the 1970s onward (Price et al., 1998; Kang et al., 2015; Hoell et al., 2017; Dasari et al., 2021). Although precipitation tends to increase during El Niño and to decrease during La Niña, several studies suggested that the strength of this linkage varies during the wet season, with the strongest influence during autumn, a weak influence during winter, and a modest influence in spring (Mariotti et al., 2005; Mariotti, 2007; Bahrami et al., 2021; Hochman et al., 2024; Alizadeh and Mousavizadeh, 2025; Nuroozi et al., 2025).

Previous studies have suggested various mechanisms by which forcing in tropical Pacific Sea surface temperatures (SSTs) can modulate precipitation variability over the remote Middle East (Barlow et al., 2016). Anomalies in tropical convection, linked to SST variability in the tropical central Pacific, generate a barotropic Rossby wave train that propagates eastward from the Pacific towards the Middle East via the extratropics (Kumar et al., 2016; Hoell et al., 2013, 2015a; Abid et al., 2020). During El Niño, these circulation patterns are accompanied by a cyclonic anomaly over the Middle East, enhancing precipitation through the advection of warm, moist air and large-scale ascent (Chakraborty et al., 2006; Mariotti, 2007; Hoell et al., 2018). Conversely, during La Niña, an anticyclonic circulation anomaly emerges over the region, reducing precipitation through cold air advection, weakened moisture transport, and large-scale descent (Abid et al., 2016; Hoell et al., 2014a, b, 2018). Furthermore, El Niño conditions are associated with an equatorward shift of the subtropical jet over North Africa and the Middle East, leading to increased westerly winds and storm activity over the region, and thereby increasing precipitation (Kumar et al., 2016; Alizadeh-Choobari et al., 2018; Sandeep and Ajayamohan, 2018; Hoell et al., 2024). Conversely, under La Niña conditions, a poleward displacement of the subtropical jet leads to decreased westerly winds, storms, and precipitation. Other studies have suggested that tropical convection anomalies over the Indo-Pacific modify the tropical circulation through a Rossby wave response in the form of an upper-tropospheric warm structure extending northwestward toward the Middle East region, consistent with the Gill-Matsuno model (Barlow et al., 2002; Hoell et al., 2012, 2015a, 2018). In summary, previous studies have suggested that Indo-Pacific SST variability can influence Middle East precipitation through three atmospheric pathways: a barotropic Rossby wave response moving eastward via the midlatitudes, a meridional displacement of the subtropical jet, and a baroclinic response in the tropical circulation to tropical convection extending westward in the form of a Gill-Matsuno gyre.

Previous work has also identified an influence of Indian Ocean SSTs on Middle East precipitation. Variability in Indian Ocean SSTs is often expressed by the Indian Ocean Dipole, with positive SST anomalies in the west and negative anomalies in the east during its positive phase, and vice versa during its negative phase (Saji et al., 1999). Over the past four decades, the positive (negative) phase of the Indian Ocean Dipole has been associated with enhanced (reduced) precipitation across the Middle East (Athar, 2015; Pourasghar et al., 2019; Tuel et al., 2022; Hochman et al., 2024). Moreover, it has been argued that El Niño (La Niña) events co-occurring with a positive (negative) Indian Ocean Dipole have an amplifying effect on precipitation anomalies (Chakraborty et al., 2006; Abid et al., 2020; Hochman and Gildor, 2024; Hoell et al., 2025). It is important to note that ENSO and the Indian Ocean Dipole are interconnected, but that they exhibit different seasonality. El Niño tends to coincide with the positive phase of the Indian Ocean Dipole, while La Niña aligns with its negative phase (Schott et al., 2009; Stuecker et al., 2017). ENSO usually peaks during boreal winter, whereas the Indian Ocean Dipole is mainly active during boreal summer and autumn (Schott et al., 2009). This suggests that the Indian Ocean basin can act as a pathway through which ENSO influences precipitation variability in the Middle East during specific seasons (Abid et al., 2020).

While the relationship between ENSO and Middle East precipitation variability has been widely investigated, previous studies tend to focus on (1) cool season averages or single seasons only, (2) specific areas or countries within the broader Middle East, (3) correlation with indices that represent ENSO variability without accounting for the asymmetric influences of ENSO warm and cold phases, and (4) only one of the mechanistic pathways through which ENSO can modulate Middle East precipitation variability. This study presents an integrated perspective on the influence of ENSO on Middle East precipitation variability at local and spatially aggregated scales during the cool-season and its individual months. Furthermore, this study explores the atmospheric processes through which ENSO shapes precipitation variability in the Middle East by quantifying anomalies in large-scale circulation patterns and synoptic-scale weather systems. Specifically, we extend previous studies by diagnosing the latitudinal position of the subtropical jet stream and its linkage with Rossby wave breaking and by quantifying thermodynamic characteristics of the atmosphere such as atmospheric moisture content and transport and tropospheric instability. Finally, we synthesize various mechanisms by which tropical Pacific SST variability shapes regional circulation patterns and hydroclimatic conditions over the Middle East.

This study is structured as follows. Section 2 details the methodology, diagnostics, and data used. Section 3 quantifies the influence of the warm, neutral, and cold ENSO phases on Middle East precipitation variability at local and regional scales and during the extended cool season and its individual months. Section 4 explores the influence of ENSO on the subtropical jet and extratropical Rossby wave breaking. Section 5 examines the effect of ENSO on tropospheric moisture and instability over the Middle East. Section 6 discusses different mechanisms and pathways from a global perspective. Finally, Sect. 7 summarizes the main findings and limitations of the study.

Figure 2 presents the spatial distribution, time evolution, and statistical relationship between cool-season Middle Eastern precipitation and ENSO as represented by the Nino3.4 index. Consistent with the consensus in the literature, we observe an increase in precipitation during El Niño and a decrease during La Niña. During El Niño events, precipitation exhibits a weak but significant increase over parts of the Middle East, exceeding 10 % and locally reaching up to 20 % above climatology (Fig. 2a). In contrast, La Niña events show a moderate and significant decrease in precipitation falling below 20 % of climatology across much of the region and decreasing by more than 35 % over the Sinai Peninsula, the southern parts of the Arabian Peninsula, and the southern coasts of Iran and Pakistan (Fig. 2c). During the neutral ENSO phase, precipitation does not deviate significantly from climatology apart from the northern part of the Red Sea and surrounding land areas, where precipitation increases by up to 35 % (Fig. 2b).

Consistent with the spatial distribution, the timeseries of spatially aggregated cool-season precipitation generally show positive anomalies during El Niño events and negative anomalies during La Niña events (Fig. 2d). Particularly striking are the multi-year La Niña events during 1998/1999–2000/2001 and 2007/2008–2011/2012, both of which were associated with severe drought (Barlow et al., 2002; Hoell et al., 2014a). The timeseries also indicate some atypical seasons, such as a decrease in precipitation during the El-Niño seasons of 1986/1988, 2002/2003, 2009/2010, and 2014/2015, and an increase in precipitation during the La Niña seasons of 1984/1985 and 1995/1996. These deviations from the commonly observed ENSO – Middle East precipitation patterns have previously been linked to variations in SST gradients between the western and central Pacific and to the extension of Pacific SST anomalies into the eastern Indian Ocean (Hoell and Funk, 2013; Hoell et al., 2014b). The timeseries also hint at interdecadal variability in Middle East precipitation, typically showing wet conditions during the period from 1979 to 1998, followed by two decades of dry conditions from 1999 to 2017. Earlier studies have suggested a linkage with multi-decadal variability in both the Pacific and Atlantic Oceans (Kang et al., 2015; Hoell et al., 2015b; Dasari et al., 2021; Rezaei, 2023; Saharwardi et al., 2024), although their respective contributions and underlying mechanisms remain unclear.

The statistical distribution of spatially aggregated precipitation differs substantially among ENSO warm, neutral, and cold phases (Fig. 2e). Relative to climatology, the spatially aggregated mean precipitation is significantly higher during El Niño and significantly lower during La Niña. Consistent with these differences between ENSO warm and cold phases, the interquartile range of precipitation during El Niño lies entirely above the interquartile range of precipitation during La Niña (red and blue boxes in Fig. 2e). Additionally, the relationship between cool-season precipitation and ENSO exhibits substantial asymmetry (Fig. 2e, f). Mean anomalies in spatially aggregated precipitation from the multi-year mean (99.3) are substantially larger during La Niña (-19 mm) than during El Nino (+14 mm; Fig. 2e). Linear regression of cool-season precipitation as a function of the Nino3.4 index shows a change in precipitation per degree of central Pacific SST anomalies (i.e., the slope) of approximately 25.5 and 9.1 mm °C⁻¹ for Nino3.4 index values below -0.5 and above 0.5, respectively, a nearly threefold difference (Fig. 2f). The slope in precipitation to ENSO forcing is significant for Nino3.4 index values below -0.5 with a p-value < 0.11 and non-significant for Nino3.4 index values above 0.5 (p-value = 0.41). The Pearson correlation between the Nino3.4 index and spatially aggregated precipitation amounts is 0.44 and 0.24 for Nino3.4 index values below -0.5 and above 0.5, respectively. These computations demonstrate that the Middle Eastern precipitation response to ENSO forcing is stronger for La Niña than for El Niño.

On monthly timescales, Middle East precipitation shows a similar relationship with ENSO as in the cool-season average, albeit with some inter-monthly variations (Figs. 3 and A1). During El Niño conditions, the spatially aggregated precipitation increases substantially during autumn and modestly during spring compared with neutral ENSO and La Niña conditions (Fig. 3a). The spatially aggregated precipitation average is more than doubled in October and approximately one-and-a-half times larger in November compared with neutral ENSO conditions. In the spring months (March, April, and May), precipitation also enhances under El Niño conditions compared with neutral ENSO conditions with a significant increase in the spatially aggregated mean for the month of May only. Note that the precipitation increase under El Niño conditions is not only apparent from the mean, but also from the precipitation distribution in general, showing substantial increases for the 25th, 50th, and 75th percentiles (Fig. 3a). Conversely, La Niña conditions consistently exhibit a substantial reduction in precipitation compared with El Niño and neutral ENSO conditions throughout the cool season with significant decreases in the spatially aggregated mean during October, December, March and May at the p<0.05 level and for November and April at the p<0.1 level (Fig. 3c). During the core winter months of January and February, precipitation tends to be higher under neutral ENSO conditions compared to both El Niño and La Niña conditions, with a significant increase in the spatially aggregated mean for the month of January at p<0.1 (Fig. 3b). Consistent with this notion, Hoell et al. (2015a) documented different correlation patterns of hydroclimate conditions over Southwest Asia with tropical Pacific SST anomalies for the month of January compared with other cool-season months.

In summary, our findings show that ENSO has an asymmetric influence on Middle East precipitation with a stronger decrease during La Niña and a somewhat milder increase during El Niño (cf. Hoell et al., 2018). This relationship varies strongly within the cool season with, under El Niño, the largest precipitation increases during autumn (cf. Mariotti et al., 2005; Mariotti, 2007; Bahrami et al., 2021; Hochman et al., 2024; Alizadeh and Mousavizadeh, 2025) and modest increases during spring, contrasting with a persistent precipitation decrease under La Nina conditions throughout the cool season. The asymmetric influence and intraseasonal variations are largely missed when considering cool-season averages or simple correlations with ENSO indices, and thus must be considered when assessing ENSO-related precipitation variability and predicting hydrometeorological hazards in the region. Note that the influence of ENSO on Middle Eastern precipitation variability is susceptible to decadal variability, and thus, does not ensure a similar relationship before and after our period of consideration nor a stationary relationship during our period of consideration (1979–2020).

Midlatitude forcing plays a key role in the formation of precipitation across the Middle East during the cool season. Extratropical Rossby waves propagate and break towards lower latitudes, resulting in the formation of upper-level troughs and often cut-off lows (Kahana et al., 2002; Kumar et al., 2015; de Vries et al., 2016, 2018; Tuel et al., 2022; Francis et al., 2025). These upper-level cyclonic circulation patterns can induce cyclogenesis near the surface, intense atmospheric moisture transport from southerly directions, and large-scale ascent, leading to tropospheric conditions conducive to the occurrence of deep moist convection. In this section, we analyze the anomalies in the subtropical jet and Rossby wave breaking occurrences during precipitation and dry-day events at daily timescales to provide a reference for interpreting similar anomalies during ENSO warm and cold phases at seasonal to monthly timescales.

Figure 4 shows anomalies in zonal wind and relative anomalies in Rossby wave breaking frequencies based on precipitation and dry-day events in the Middle East. Rossby wave breaking frequencies are here defined by the counts of stratospheric PV structures summed across isentropic surfaces from 300 to 350 K with 5 K intervals. Anomalies in zonal winds demonstrate that precipitation events are accompanied by increased zonal winds at the equatorward flank of the subtropical jet over North Africa and the Middle East and decreased zonal winds at its poleward flank, suggesting an equatorward shift of the subtropical jet (Fig. 4a, b). Rossby wave breaking frequencies more than double, evident from relative anomalies in stratospheric PV structure frequencies exceeding 100 % above climatology, primarily on the poleward flank of the subtropical jet (between 20–33° N) and beneath the jet core (between 800 and 200 hPa), confirming the central role of Rossby wave breaking in Middle Eastern precipitation (Fig. 4a, b). In contrast, during dry-day events, the subtropical jet shifts poleward and Rossby wave breaking frequencies decrease at the poleward edge and beneath the subtropical jet (Fig. 4c, d). The southwest-northeasterly orientation of PV structure frequency anomalies during both precipitation and dry-day events is suggestive of anticyclonic Rossby wave breaking imposing an upper-tropospheric trough and ridge over the region during wet and dry conditions, respectively. Furthermore, during both precipitation and dry-day events, anomalies in PV structure frequencies of opposite sign emerge farther upstream and downstream of the largest anomalies over the eastern Mediterranean, reflecting the typical structure of Rossby wave packets with multiple neighboring troughs and ridges embedded within wave packets.

Consistent with previous studies (Kumar et al., 2016; Alizadeh-Choobari et al., 2018; Sandeep and Ajayamohan, 2018), Fig. 5 shows that El Niño conditions are associated with an equatorward shift of the subtropical jet, whereas La Niña conditions are associated with a poleward shift of the jet. During El Niño events, positive zonal wind anomalies appear on the equatorward flank of the jet, peaking above its core near 150 hPa, while negative zonal wind anomalies occur on the poleward flank of the jet (Fig. 5a, b). These changes in the background zonal wind go along with increased frequencies in stratospheric PV structures exceeding 10 % on the poleward flank and beneath the subtropical jet core (Fig. 5a, b). Conversely, La Niña events demonstrate opposite patterns with negative zonal wind anomalies on the equatorward side and positive anomalies on the poleward side of the jet, going along with a reduction of more than 10 % in stratospheric PV structure frequencies poleward and beneath the subtropical jet (Fig. 5c, d). In essence, composites of anomalies in the subtropical jet and Rossby wave breaking occurrences during ENSO warm and cold phases resemble the patterns observed during precipitation and dry-day events. This suggests that El Niño events bring about wet conditions through an equatorward displacement of the subtropical jet accompanied by increased Rossby wave breaking occurrences, while La Niña events exhibit dry conditions linked to a poleward displaced jet associated with decreased Rossby wave breaking occurrences. Figure 5 also shows anomalies in stratospheric PV structure frequencies with opposite sign at the equatorward flank of the subtropical jet (between 10–20° N), that is, negative anomalies during El Niño and positive anomalies during La Niña. However, given that these anomalies arise outside regions of stratospheric PV structure imprints during precipitation and dry-day events, shown by purple and green contours in Fig. 5, respectively, we consider these anomalies irrelevant for understanding precipitation anomalies during ENSO warm and cold phases.

On monthly timescales, similar patterns emerge as during the cool season average, displaying an equatorward (poleward) displacement of the subtropical jet during El Niño (La Niña) conditions (Figs. A2, A3). Increased (decreased) Rossby wave breaking occurs beneath and poleward of the subtropical jet during El Niño (La Niña) conditions, primarily during October, November, and December and to a lesser extent in March and May. To further quantify the relationship between ENSO and the large-scale circulation over the Middle East, we evaluate the distribution of the latitudinal position of the jet stream during each month and ENSO phase (Fig. 6). The multi-year monthly mean jet position during different ENSO phases, depicted by the dots, generally aligns with the monthly distribution of anomalies in precipitation and Rossby wave breaking occurrences. Under El Niño conditions, the jet is displaced southward in October (p<0.05), November (p<0.1), December and April (non-significant) compared with neutral ENSO conditions (Fig. 6a), consistent with increased precipitation and wave breaking frequencies during these months (Figs. 3a, A2, A3). Under La Niña conditions, the jet is displaced poleward in January, February, and May (p<0.05) and in November and March (non-significant), compared with neutral ENSO conditions (Fig. 6c), consistent with decreased precipitation and wave breaking frequencies during these months (Figs. 3c, A2, A3). Some months deviate from these patterns, showing jet displacements that are inconsistent with anomalies in precipitation and/or Rossby wave breaking frequencies. For example, under El Niño conditions, the jet shows a significant southward displacement in February (Fig. 6a), which contrasts with the non-significant precipitation decrease during that month relative to neutral ENSO conditions (Fig. 3a). Similarly, under La Niña conditions, the mean jet position in December is very similar to that under neutral ENSO conditions (Fig. 6c), despite the significant decrease in precipitation and Rossby wave breaking frequencies compared with neutral ENSO conditions during this month (Figs. 3c, A2, and A3).

In summary, apart from some deviations in specific months, the atmospheric circulation composites generally show that El Niño conditions are associated with a southward displaced jet and increased wave breaking frequencies at its poleward flank and beneath its core, while La Niña conditions are associated with a poleward displaced jet and decreased wave breaking frequencies. The similarity between the ENSO-based composites and those based on precipitation and dry-day events suggests that El Niño and La Niña events induce wet and dry conditions, respectively, through latitudinal shifts in the jet stream accompanied by anomalies in Rossby wave breaking frequencies.

Precipitation in the Middle East is associated with intrusions of warm and moist air masses from southerly directions that support the formation of deep moist convection. Previous studies have shown that ENSO variability is linked to anomalies in atmospheric moisture content and transport over the Middle East (Chakraborty et al., 2006; Mariotti et al., 2005, Mariotti, 2007; Hochman and Gildor, 2024). Consistent with these findings, Fig. 7a shows increased atmospheric moisture content and strengthened moisture transport from southwesterly directions during El Niño conditions. It is important to note that climatological moisture transport over the Middle East is primarily from southwesterly directions (not shown). Conversely, under La Niña conditions, atmospheric moisture content decreases, and anomalies in atmospheric moisture transport have northeasterly directions, suggesting reduced moisture transport into the region (Fig. 7c). These anomalies emerge at the southeastern flanks of mid-tropospheric cyclonic and anticyclonic circulation anomalies, pointing to ENSO forcing on midlatitude circulation patterns through which El Niño and La Niña conditions strengthen and weaken moisture transport into the Middle East, respectively. The alignment of anomalous IVT vectors with 500 hPa geopotential height anomaly contours suggests a barotropic Rossby wave response to tropical Pacific SST forcing. However, farther southeast from the region of interest, anomalous IVT vectors display an approximately perpendicular or even opposing alignment with 500 hPa geopotential height anomalies, suggesting a baroclinic response of tropospheric circulation to ENSO variability. We will return to this aspect in Sect. 6.

On monthly timescales, anomalies in atmospheric moisture content and transport show patterns very similar to the cool season average (Fig. A4), while monthly variations in these patterns align with the monthly precipitation anomalies (Fig. A1). Spatially aggregated total column water content follows a strong seasonal cycle with the highest values in summer and the lowest values during winter (Fig. 8). Despite the greatest availability of moisture in summer, the region remains dry due to large-scale subsidence over the broader Middle East, which suppresses upward motion (Tyrlis et al., 2013). During autumn, westerly winds and the subtropical jet migrate southward, along with the decay of the South Asian summer monsoon circulation, allowing extratropical disturbances to reach the region. The availability of both dynamically forced ascent and atmospheric moisture during the transition seasons and in winter facilitates the occurrence of precipitation during this period of the year. Spatially aggregated mean values of atmospheric moisture content show a significant increase under El Niño conditions during October (p<0.05), November (p<0.1), and May (p<0.05) (Fig. 8a). Conversely, La Niña conditions show decreases during all cool season months compared with neutral ENSO condition with significant departures from climatology (p < 0.05) during October–December and March–May (Fig. 8c). In January, moisture content is significantly increased for neutral ENSO conditions (Fig. 8b), consistent with the precipitation increase during that month and ENSO phase (Fig. 3b).

To provide a complementary perspective on how ENSO shapes a tropospheric environment conducive to deep moist convection, we evaluate the spatial distribution and spatially aggregated mean values of tropospheric instability. During the cool season, Imax⁡ exhibits patterns similar to those in TCW with significant positive anomalies across the Middle East during El Niño and significant negative anomalies during La Niña (Fig. 9). Anomalies are particularly large over the Arabian Peninsula and along the southern flanks of the Zagros Mountains. Anomalies of opposite sign emerge over the Red Sea and Persian Gulf, that is, negative anomalies during El Niño and positive anomalies during La Niña, most likely resulting from weak negative and positive SST anomalies over these waters during El Niño and La Niña conditions, respectively (not shown). On monthly timescales, Imax⁡ shows a strong seasonal cycle with the largest values during summer and the lowest values during winter (Fig. 10). Monthly anomalies in Imax⁡ for ENSO conditions follow anomalies in TCW and precipitation (Fig. A5). Under El Niño conditions, spatially aggregated mean values of Imax⁡ show positive anomalies during October, November, December and April being significant (p<0.05) for October only (Fig. 10a). During La Niña conditions, negative anomalies persist throughout all cool season months with significant departures from climatology in October and May at p<0.05 and March and April at p<0.1 (Fig. 10c).

We now have a clearer understanding of how precipitation and regional circulation patterns over the Middle East relate to ENSO. Still, the question remains how remote SST forcing in the tropical Indo-Pacific brings about these changes in atmospheric circulation and surface weather over a region on the opposite side of the globe. As articulated in the introduction, previous studies have proposed various mechanisms and pathways through which ENSO influences the hydroclimate of the Middle East (Barlow et al., 2016). Here we argue, based on physical reasoning and global composites, that all three previously proposed atmospheric mechanisms hold validity, and that co-varying atmosphere-ocean interactions over the Indian Ocean are likely responsible for the seasonally varying influence of ENSO on Middle Eastern precipitation.

Seager et al. (2003, 2005) have shown that hemispheric symmetric variability in precipitation at interannual timescales can be attributed to ENSO (see also Fig. 11a, d). During El Niño conditions, increased heating of the tropical atmosphere is associated with an equatorward shift and strengthening of the subtropical jet in the global zonal mean along with increased baroclinicity at lower latitudes (Seager et al., 2003). While their focus was primarily on the eastward extension of the subtropical jet over the East Pacific, leading to a regional increase in precipitation over the West Coast of North America and the southern United States (Seager et al., 2005), the equatorward shift in the subtropical jet also manifests over North Africa, the Middle East, and South Asia (Fig. 11c). This equatorward shift goes along with increased Rossby wave breaking frequencies on the poleward flank of the jet over northwest Africa, the Middle East, and southwest and central South Asia. This notion is consistent with previous studies reporting an equatorward shift of the subtropical jet and increased (extreme) precipitation at the western flanks of the Himalayas (Cannon et al., 2017; Bharati et al., 2025; Hunt et al., 2025). Conversely, during La Niña conditions, zonal winds weaken at the southern flank of the subtropical jet and strengthen at its poleward side across a region spanning from the East Atlantic through North Africa, the Middle East, South Asia, and reaching the west Pacific coast (Fig. 11f). Rossby wave breaking occurrences are significantly reduced along the entire poleward flank of the subtropical jet. Therefore, ENSO forcing induces a zonally symmetric response in the circulation, manifesting as a meridional shift of the subtropical jet and modifications in Rossby wave breaking frequencies at its poleward flank, bringing about precipitation anomalies across a larger region encompassing the Middle East and southwest and central Asia (Fig. 11a, d).

Seager et al. (2005) also recognized a zonally asymmetric component of the precipitation response to ENSO, which is generally explained by a Rossby wave response to anomalous tropical Pacific SSTs and convection (Wallace and Gutzler, 1981; Hoskins and Ambrizzi, 1993). These stationary Rossby wave patterns emanate from the tropical central Pacific and arch in an eastward direction across North America, the North Atlantic, and reach North Africa and Europe (Fig. 11b, e). During El Niño events, this Rossby wave response appears as an anticyclonic anomaly over northwest Africa and southwestern Europe and a cyclonic anomaly over the Middle East (Fig. 11b), offsetting and strengthening the precipitation response to the zonally symmetric shift of the jet stream and wave breaking anomalies, respectively. During La Niña events, the signs of these anomalies reverse, with cyclonic and anticyclonic circulation patterns over northwest Africa-southwestern Europe and the Middle East, respectively (Fig. 11e). These patterns correspond to previously documented responses of the atmospheric circulation to tropical convection anomalies in the Indo-Pacific at interannual timescales, manifesting as an equivalent barotropic Rossby wave response derived from their consistent vertical structure (Hoell et al., 2012, 2013, 2015a; Kumar et al., 2016). Additionally, several studies have also identified a baroclinic response in the tropical circulation over the Indian Ocean–South Asia sector to tropical convection anomalies over the Warm Pool, taking the form of a Gill-Matsuno-gyre (Barlow et al., 2002; Hoell et al., 2012, 2015a). The Gill-Matsuno response explains that anomalous tropical heating (cooling) leads to a Rossby wave response, interacting with the mean flow, in the form of an upper-tropospheric anticyclonic (cyclonic) circulation anomaly to the northwest and southwest, accompanied by a lower-tropospheric cyclonic (anticyclonic) circulation anomaly (Matsuno, 1966; Gill, 1980). Accordingly, Fig. 11b shows an upper-tropospheric cyclonic circulation anomaly centered over southeast Asia (100° E and 30° N) to the northwest of reduced tropical convection over the Warm Pool. This cyclone extends westward across South Asia and the Middle East. In the lower troposphere, an anticyclonic circulation anomaly is evident based on the clockwise orientation of anomalous atmospheric moisture transport over the Indian Ocean and South Asia, including anomalous southwesterly IVT over the Middle East, implying strengthened atmospheric moisture transport into the region. The reversed pattern manifests during La Niña, with an upper-troposphere anticyclonic anomaly emerging over southeast Asia to the northwest of enhanced tropical convection over the Warm Pool. This circulation anomaly extends westward across South Asia and the Middle East, accompanied by an anomalous counterclockwise circulation in the lower troposphere, setting up anomalous northeasterly moisture transport over the Middle East, and thus reducing moisture transport into the region (Fig. 11e). The circulation response over the Middle East to Indo-Pacific tropical convection anomalies is consistent with the monsoon-desert mechanism of Rodwell and Hoskins (1996, 2001). According to this mechanism, reduced (enhanced) tropical convection sets up west-to-east upsloping (down sloping) isentropic surfaces facilitating isentropic upward-moving (down-gliding) air masses, resulting in large-scale ascent (subsidence), favoring (suppressing) precipitation formation (Barlow et al., 2005, 2007). This mechanism is evident from widespread mid-tropospheric ascent (descent) over the broader Middle East region in Fig. 9a (Fig. 9c).

To summarize, bringing together the findings from previous research and supported by composites in Fig. 11, we postulate that ENSO modulates precipitation variability in the Middle East through a combination of at least three distinct atmospheric pathways (Fig. 13): (1) a zonally symmetric latitudinal shift in the subtropical jet stream (Kumar et al., 2016; Alizadeh-Choobari et al., 2018; Sandeep and Ajayamohan, 2018; Hoell et al., 2024) alongside changes in Rossby wave breaking occurrences at its poleward flank, (2) a barotropic Rossby wave pattern arching from the central tropical Pacific eastward across North America, the Atlantic, and the European–North Africa region (Kumar et al., 2016; Hoell et al., 2013, 2015a; Abid et al., 2020), and (3) a baroclinic response in the tropical circulation extending westwards from the Warm Pool across the Indian Ocean-South Asia sector into the Middle East, consistent with a Gill-Matsuno-type response (Barlow et al., 2002; Hoell et al., 2012, 2015a). While these mechanisms contribute to precipitation anomalies during the cool season, the inter-monthly variations of precipitation to ENSO forcing remain enigmatic. To investigate this aspect further, we focus our attention on the co-varying state of the atmosphere over the Indian Ocean.

Figure 11b, e not only shows outgoing longwave radiation anomalies over the central and western Pacific, but also over the western Indian Ocean. To further explore this signal, we construct Hovmöller diagrams of anomalies in monthly mean outgoing longwave radiation across the equator as a function of longitude and season under El Niño and La Niña conditions (Fig. 12). Besides the expected dipole in tropical convection over the central and western Pacific, which persists throughout the year, we also notice a seasonally asymmetric response in tropical convection over the western Indian Ocean. Specifically, under El Niño conditions, we observe significantly increased tropical convection over the western Indian Ocean (40–70° E) from October to December and weak anomalies of the same sign in May. Conversely, under La Niña conditions, tropical convection decreases significantly over the western Indian Ocean from October to December and again in April and May. In essence, ENSO-driven tropical convection anomalies over the western Indian Ocean follow the same pattern as the seasonally varying and asymmetric response of precipitation in the Middle East to ENSO (Fig. 3). This suggests that ENSO, beyond the three atmospheric pathways, also influences the Middle East hydroclimate through its effect on the Indian Ocean and atmospheric state (cf. Abid et al., 2020; Hochman and Gildor, 2024).

This study provides a comprehensive analysis of how ENSO influences precipitation variability in the arid Middle East during the extended cool season (October–May), when the region receives most of its annual rainfall. Understanding how tropical Indo-Pacific SST variability modulates rainfall over the water-scarce Middle East is highly relevant to society, given the profound impacts of drought and flooding on the population, economy, and agriculture. While several of the findings presented in this study have been documented in previous studies, we present here, for the first time, an integrated perspective on the precipitation response to ENSO forcing and the underlying atmospheric drivers. We extend previous work by quantifying the influence of ENSO on precipitation variability at local and regional scales during the cool season and individual months. Furthermore, we employ various diagnostics to quantify the influence of ENSO on the latitudinal position of the subtropical jet stream and the occurrence of Rossby wave breaking and on thermodynamic characteristics such as tropospheric moisture content, moisture transport, and instability. This section summarizes how ENSO forcing shapes the Middle Eastern hydroclimate by synthesizing knowledge from existing studies, complemented with the new findings presented in this study.

Consistent with existing literature (e.g., Price et al., 1998; Nazemosadat and Cordery, 2000; Nazemosadat and Ghasemi, 2004; Mariotti, 2007; Athar, 2015; Kumar et al., 2016; Alizadeh and Mousavizadeh, 2025), we find that El Niño is generally associated with increased precipitation over the Middle East and La Niña with decreased precipitation. The precipitation response to ENSO is asymmetric and varies within the cool season. Anomalies in local and spatially aggregated precipitation are larger during La Niña than during El Niño. Likewise, the slope in spatially aggregated precipitation as a function of the Niño3.4 index is nearly three times larger during La Niña than during El Niño. At intraseasonal scales, precipitation is strongly increased during autumn and moderately enhanced in spring under El Niño conditions. Conversely, during La Niña conditions, precipitation decreases throughout the cool season with significant reductions relative to climatology in autumn and spring. In the core winter months of January and February, precipitation increases under neutral ENSO conditions. The influence of ENSO on Middle East precipitation variability is directly relevant to hydrometeorological hazards. Consecutive La Niña events have been linked to multi-year droughts during the periods of 1999–2001 and 2007–2011 (Barlow et al., 2002; Hoell et al., 2014a). Similarly, El Niño conditions coincided with several devastating floods in the Levant in autumn of 1987, 1994, and 1997 and in the United Arab Emirates in spring 2024 (Krichak et al., 2012; de Vries et al., 2013; Kadhum et al., 2022; Francis et al., 2025; Hussein et al., 2025), corresponding to regions and seasons with increased precipitation under El Niño conditions (Fig. A1).

These precipitation anomalies during the ENSO warm and cold phases are associated with consistent changes in regional circulation patterns over the Middle East. During El Niño conditions, the subtropical jet shifts equatorward, as pointed out by Kumar et al. (2016), Alizadeh-Choobari et al. (2018), and Sandeep and Ajayamohan (2018), accompanied by increased Rossby wave breaking frequencies at the poleward flank and beneath the jet. Consistent with this pattern, a mid-tropospheric cyclonic circulation anomaly dominates over the Middle East along with elevated atmospheric moisture content and strengthened atmospheric moisture transport at its southeastern flank. Likewise, tropospheric conditions are more prone to deep moist convection, facilitated by anomalous warm and moist air advection in the lower troposphere and anomalous cold air in the upper troposphere. Under La Niña conditions, these patterns are reversed. The subtropical jet shifts poleward (cf. Hoell et al., 2024) with a reduction in Rossby wave breaking frequencies at its poleward flank. A mid-tropospheric anticyclonic circulation anomaly emerges, associated with reduced atmospheric moisture content and conditional instability over the Middle East along with weakened atmospheric moisture transport into the region.

These regional precipitation and atmospheric circulation patterns over the Middle East are the combined result of three atmospheric mechanisms (Fig. 13). From a zonal mean perspective, El Niño conditions are associated with an equatorward shift and strengthening of the subtropical jet (Seager et al., 2003, 2005) accompanied by increased Rossby wave breaking at its poleward fringe across North Africa, the Middle East, and southwest-central Asia (Fig. 13a). Anomalous heating in the central Pacific sets up a well-known stationary barotropic Rossby wave pattern arching across North America, the North Atlantic, Europe and North Africa, and culminating in a cyclonic circulation anomaly over the Middle East (Hoell et al., 2013, 2015a; Kumar et al., 2016). This cyclonic circulation anomaly is merging with a strong upper-tropospheric cyclonic anomaly centered over southeast Asia, which can be understood as a Gill-Matsuno-type response to reduced tropical convection over the tropical Warm Pool, extending westward over South Asia (Barlow et al., 2002; Hoell et al., 2012, 2015a). Consistent with this baroclinic circulation response, an anticyclonic circulation anomaly in the lower troposphere feeds increased warm, moist air masses into transient storms passing by from the west (Mariotti et al., 2005; Mariotti, 2007). La Niña conditions generate opposite circulation patterns, thereby reducing precipitation over the Middle East (Fig. 13b). Besides these three atmospheric pathways, ENSO likely also influences precipitation in the Middle East via its effect on Indian Ocean SSTs and atmospheric conditions (Abid et al., 2020; Hochman and Gildor, 2024). Hence, the distinct seasonal cycle of the Indian Ocean Dipole, mainly active during boreal summer and autumn, may explain why the influence of ENSO on Middle East precipitation varies within the cool season. While our analysis brings together various mechanisms through which tropical Pacific SST forcing modulates precipitation variability over the Middle East, future research may further quantify and disentangle the relative importance of these mechanisms, for example, based on numerical model experiments. Based on our findings, we anticipate that the relative importance of these mechanisms may vary across ENSO phases and seasons.

The findings in this study are subject to several limitations. First, the precipitation dataset used inherits the limitations of the underlying data sources, although qualitatively similar results based on ERA5 precipitation strengthen the confidence in our findings (not shown). Second, our study covers a 42-year period, which inherently limits the sample size of ENSO events and ENSO-like conditions. Third, our analysis does not account for ENSO diversity, such as variations in west-central Pacific SST gradients or the eastward extent of Pacific SST anomalies into the eastern Indian Ocean (Hoell and Funk, 2013; Hoell et al., 2014b, 2017, 2018), which may modify the Middle Eastern precipitation response. Fourth, our study does not account for non-stationarity in the ENSO-Middle East precipitation relationship, which has been shown to vary across the 20th century (Price et al., 1998; Kang et al., 2015; Hoell et al., 2017; Dasari et al., 2021). Nonetheless, our study provides clear evidence of how ENSO modulates precipitation variability over the Middle East and the underlying atmospheric mechanisms bringing about these precipitation anomalies. Improved process understanding of how tropical Pacific SST variability shapes hydroclimatic conditions in the Middle East is directly relevant to seasonal prediction of precipitation over this water-scarce region, with implications for drought and flood preparedness, as well as for identifying climate model biases and reducing uncertainty in climate projections (Barlow et al., 2021).