What Controls the Interannual Variability of the Boreal Winter Atmospheric River Activities over the Northern Hemisphere?

Weiming Ma aDepartment of Atmospheric and Oceanic Sciences, University of California, Los Angeles, Los Angeles, California

Search for other papers by Weiming Ma in
Current site
Google Scholar
PubMed
Close
and
Gang Chen aDepartment of Atmospheric and Oceanic Sciences, University of California, Los Angeles, Los Angeles, California

Search for other papers by Gang Chen in
Current site
Google Scholar
PubMed
Close
Free access

Abstract

Interannual variability of the winter atmospheric river (AR) activities over the Northern Hemisphere is investigated. The leading modes of AR variability over the North Pacific and North Atlantic are first identified and characterized. Over the Pacific, the first mode is characterized by a dipole structure with enhanced AR frequency along the AR peak region at about 30°N and reduced AR frequency farther north. The second mode exhibits a tripole structure with a narrow band of positive AR anomalies at about 30°N and sandwiched by negative anomalies. Over the Atlantic, the first mode exhibits an equatorward shift of the ARs with positive anomalies and negative anomalies located on the equatorward and poleward side of the AR peak region at about 40°N, respectively. The second mode is associated with the strengthening and eastward extension of the AR peak region, which is sandwiched by negative anomalies. A large ensemble of atmospheric global climate models from phase 6 of the Coupled Model Intercomparison Project (CMIP6), which shows high skill in simulating these modes, is then used to quantify the roles of sea surface temperature (SST) forcing versus internal atmospheric variability in driving the formation of these modes. Results show that SST forcing explains about half of the variance for the Pacific leading modes, while that number drops to about a quarter for the Atlantic leading modes, suggesting higher predictability for the Pacific AR variability. An additional ensemble driven only by observed tropical SST is further utilized to demonstrate the more important role that tropical SST plays in controlling the Pacific AR variability while both tropical and extratropical SST exert comparable influences on the Atlantic AR variability.

© 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Weiming Ma, mawei@atmos.ucla.edu

Abstract

Interannual variability of the winter atmospheric river (AR) activities over the Northern Hemisphere is investigated. The leading modes of AR variability over the North Pacific and North Atlantic are first identified and characterized. Over the Pacific, the first mode is characterized by a dipole structure with enhanced AR frequency along the AR peak region at about 30°N and reduced AR frequency farther north. The second mode exhibits a tripole structure with a narrow band of positive AR anomalies at about 30°N and sandwiched by negative anomalies. Over the Atlantic, the first mode exhibits an equatorward shift of the ARs with positive anomalies and negative anomalies located on the equatorward and poleward side of the AR peak region at about 40°N, respectively. The second mode is associated with the strengthening and eastward extension of the AR peak region, which is sandwiched by negative anomalies. A large ensemble of atmospheric global climate models from phase 6 of the Coupled Model Intercomparison Project (CMIP6), which shows high skill in simulating these modes, is then used to quantify the roles of sea surface temperature (SST) forcing versus internal atmospheric variability in driving the formation of these modes. Results show that SST forcing explains about half of the variance for the Pacific leading modes, while that number drops to about a quarter for the Atlantic leading modes, suggesting higher predictability for the Pacific AR variability. An additional ensemble driven only by observed tropical SST is further utilized to demonstrate the more important role that tropical SST plays in controlling the Pacific AR variability while both tropical and extratropical SST exert comparable influences on the Atlantic AR variability.

© 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Weiming Ma, mawei@atmos.ucla.edu

1. Introduction

Atmospheric rivers (ARs), defined as filaments of intense moisture transport in the atmosphere, exert profound impacts on the global hydrological cycle and regional weather extremes. Despite covering only about 10% of the latitude circumference in the midlatitudes at any given time, it is estimated that they are responsible for more than 90% of the poleward moisture transport (Zhu and Newell 1998). ARs are important water suppliers for many regions around the world. On the one hand, they contribute significantly to the annual total precipitation to regions, such as the North American west coast, western Europe, and southern South America (Dettinger et al. 2011; Rutz and Steenburgh 2012; Lavers and Villarini 2015; Viale et al. 2018). For example, it is estimated that California receives up to 50% of its annual total precipitation from ARs (Dettinger et al. 2011). On the other hand, they are also responsible for many weather extremes, such as extreme precipitation (Lavers and Villarini 2013; Leung and Qian 2009; Lamjiri et al. 2017; Ma et al. 2020b), extreme wind events (Waliser and Guan 2017), and rain-on-snow events (Guan et al. 2016; Bozkurt et al. 2021; McGowan et al. 2021). Recently, their presence in the tropics and the polar regions is also starting to get recognized (Woods and Caballero 2016; Mattingly et al. 2018; Nash et al. 2018; Thapa et al. 2018; Francis et al. 2020). ARs play complex roles in influencing the polar climate. Studies show that ARs can induce sea ice losses and ice sheet melting (Mattingly et al. 2018; Woods and Caballero 2016), but they can also cause anomalous snow accumulation (Gorodetskaya et al. 2014). With ARs shifting poleward in both the present and future climate (Zhang et al. 2021; Gao et al. 2016; Ma et al. 2020a), it is likely that ARs will exert stronger influences on the polar climate.

Considering the important roles that ARs play in the climate system, it is critical for both disaster preparation and water resource management to understand what controls their variability for time scales ranging from subseasonal-to-seasonal (S2S) to interannual. At the S2S time scale, the Madden–Julian oscillation (MJO) provides the dominant source of predictability for AR activities over the North Pacific (Guan et al. 2012; Zhou et al. 2021), but its role is also modulated by the quasi-biennial oscillation (QBO) (Mundhenk et al. 2018; Baggett et al. 2017). By analyzing six decades of observed data, Guirguis et al. (2019) find that Arctic Oscillation, eastern Pacific Oscillation, and western Pacific Oscillation strongly modulate subseasonal AR activities over Northern California, which tends to receive more precipitation during boreal winter when these modes are in their negative phase. At the interannual time scale, Gershunov et al. (2017) find that landfalling ARs over the western North America during boreal winter are modulated by the Pacific decadal oscillation and the “blob” of warm SST off the Pacific Northwest (Bond et al. 2015). ENSO, as the dominant mode of SST variability at the interannual time scale, has also been shown to exert strong influences on the AR landfalls along the North American west coast, with such influences being dependent on the ENSO flavors (Kim et al. 2019; Xiong and Ren 2021).

While ENSO can provide predictability for AR activities at the interannual time scale, internal atmospheric variability, which is intrinsic to the atmosphere, can also strongly modulate AR variability (Teng and Branstator 2017; Zhang et al. 2018; Dong et al. 2018; Kumar and Chen 2017; Chen and Kumar 2018; Deser et al. 2018; Cash and Burls 2019). One example is the well-known failed 2016 winter rains in Southern California. The winter of 2015/16 was associated with a major El Niño event with magnitude comparable to the 1997/98 event. Historically, strong El Niño events tend to cause wet rainy seasons over California. This led to heightened expectations that the multiyear drought that California was experiencing at that time would finally be alleviated in 2016. However, the anticipated high rainfall in Southern California did not materialize that winter, leaving the drought continuing into 2017. A later modeling study suggests that this failed rainfall is not a response to the boundary forcing and attributes the cause to internal atmospheric variability (Zhang et al. 2018). Indeed, using a large ensemble of models from the Atmospheric Model Intercomparison Project (AMIP) that participated in CMIP5 (Taylor et al. 2012), Dong et al. (2018) shows that about 80% of the interannual variability of winter precipitation in California is controlled by internal atmospheric variability. However, this view has been challenged recently. Using the newly defined ENSO Longitude Index, Patricola et al. (2020) argue that the failed 2016 winter rains in Southern California can be explained by the spatial pattern of the El Niño, rather than internal atmospheric variability. Given such complex interplay between the SST forcing and internal atmospheric variability in modulating AR and precipitation variability, a natural question to ask is: What relative roles are played by SST forcing versus internal atmospheric variability in driving the interannual AR variability?

In this study, we will address the question posed above for the winter AR variability over the Northern Hemisphere. To the best of our knowledge, only a few studies have focused on the leading modes of winter AR variability over the Northern Hemisphere so far, but the AR response to climate forcing is often dominated by the leading modes of low-frequency variability. For example, by applying empirical orthogonal function (EOF) analysis to the integrated water vapor transport (IVT) field induced by ARs, Dong et al. (2018) identifies the two leading modes of winter AR variability over the North Pacific. They show that the first mode correlates strongly with the interannual variability of extreme precipitation along the U.S. west coast. Over the North Atlantic, Li et al. (2022) identifies the first three modes of SST-forced winter AR variability using a 30-member ensemble of AMIP-style simulations. They also identify the large-scale climate modes behind these leading modes. In the first part of this study, we will first systematically identify and characterize the leading modes of interannual AR variability in the North Pacific and North Atlantic during winter in observation. A large ensemble of AMIP-type models from CMIP6 is then evaluated on their performance in simulating these modes (Eyring et al. 2016). After showing that these models are capable of producing AR climatology and variability with high fidelity, we systematically quantify the roles of SST forcing versus internal atmospheric variability in driving the interannual variability of these modes. For the forced component of AR variability, the relative roles of tropical SST versus extratropical SST are also explored.

This paper is structured as follows. In section 2, we will first describe the observed datasets and the model datasets. We will then describe the AR detection algorithm and the EOF method used to extract the leading modes of AR frequency anomalies in observation. The singular value decomposition (SVD) method used to derive the leading modes of AR anomalies for the covariance between observation and models is also described in this section. Major findings will be presented in sections 3 and 4. A brief conclusion will be provided in section 5.

2. Data and method

a. Observations and model datasets

This study focuses on the Northern Hemisphere winter season [December–February (DJF)] from 1979 to 2014, which covers the longest overlapping period between the observations and model datasets. The observed AR statistics are based on the daily mean of the 6-hourly horizontal winds and specific humidity at 1000-, 850-, 700-, and 500-mb levels from the ECMWF interim reanalysis (ERA-Interim; Dee et al. 2011) with spatial resolution of 1.5° × 1.5°. The detrended winter AR frequency anomalies are used to define the leading modes of AR variability on the interannual time scale, to be described in section 2c. Monthly sea level pressure from ERA-Interim is used to characterize the large-scale circulation associated with the leading modes of AR variability. Daily precipitation covering the period from 1996 to 2014 from the Global Precipitation Climatology Project (GPCP; Huffman et al. 2001) is used to examine the winter mean and extreme precipitation anomalies associated with the leading modes of AR variability. In this study, the winter mean precipitation is defined as the total precipitation summed over all the daily precipitation during winter. Extreme precipitation is defined as the total precipitation summed over all the winter days with daily precipitation exceeding the 95th percentile of the winter daily precipitation. To test the robustness of the results, we also examine the daily precipitation calculated from the 12-h accumulated precipitation from ERA-Interim, which covers the entire study period. The observed SST is from the Hadley Centre Sea Ice and Sea Surface temperature dataset (HadISST; Rayner et al. 2003).

AMIP models driven by observed changes in SST and sea ice are useful tools to study how boundary forcing drives the observed interannual variability of the atmosphere. To quantify the contributions of the SST/sea ice forcing versus the internal atmospheric variability in driving the interannual variability in ARs, a large ensemble of AMIP-type models that participated in CMIP6 (i.e., 30 AMIP6 models) are used in this study. Some of the models contain more than one ensemble member, but only the first member with the tag “r1i1p1f1” is used. Details on the models’ name, institution, horizontal resolution, and the number of ensemble members used for each model are described in Table S1 in the online supplemental material. Similar to ERA-Interim, daily horizontal winds and specific humidity at 1000, 850, 700, and 500 mb in the AMIP datasets are used to calculate the AR statistics.

In addition, we examine another model ensemble by including nine additional members for four of the models in the AMIP ensemble (ACCESS-ESM, CESM2, IPSL, and MIROC6), making the ensemble size of each of these four models to be 10 members. We repeat the analysis with this 40-member ensemble. The results based on this 40-member ensemble (not shown) are very similar to those based on the 30-member ensemble described in the previous paragraph. Therefore, only the results based on the 30-member ensemble are presented here (except for Figs. 6 and 7).

Two additional ensembles from the CESM1 are also employed to investigate the roles of tropical SST versus extratropical SST in driving the AR variability. Similar to the AMIP models described above, one ensemble, which consists of 10 members, is driven by the observed global SST and sea ice and is termed “Global Ocean Global Atmosphere” (GOGA). Another ensemble, which also consists of 10 members, is driven by the observed SST only in the tropical ocean from 28°S to 28°N and climatological SST poleward of 35°. A linear interpolation zone is set between 28° and 35°. This ensemble is termed “Tropical Ocean Global Atmosphere” (TOGA). Daily horizontal winds and specific humidity at the lowest model level, 850, 500, and 200 mb are used to calculate AR statistics for these two ensembles.

b. AR detection algorithm

We employ the IVT-based AR detection algorithm developed by Guan and Waliser (2015). This algorithm detects ARs globally and is shown to compare favorably with other AR detection algorithms (Guan and Waliser 2015). Many criteria, which are commonly used in other AR detection algorithms, are also employed in this algorithm, but the global algorithm used here helps identify large-scale AR activity patterns rather than local features. We will outline the major steps taken by this algorithm to detect ARs here. Readers are referred to Guan and Waliser (2015) for a more detailed description of the algorithm. Our analysis focuses on the winter mean AR frequency, which is defined as the fraction of time a grid point experiences AR condition in a given winter.

This algorithm first identifies contiguous regions of elevated IVT (“objects”) based on the seasonally and regionally dependent 85th percentile of IVT magnitude or 100 kg m−1 s−1, whichever is larger. To ensure coherence of the detected object, at least half of the grids of the object should have the IVT direction within 45° of the object mean IVT direction. Since ARs are important for poleward moisture transport, the detected object is required to exhibit a mean poleward IVT of at least 50 kg m−1 s−1. A detected AR should also be longer than 2000 km and with a length-to-width ratio greater than 2. IVT is calculated by vertically integrating the moisture flux from the lowest vertical level to the highest vertical level using the four levels available for each dataset. The IVT threshold is calculated by using the entire study period of each individual dataset. Since this algorithm detects ARs based on a percentile threshold, the results in this study are not sensitive to the different vertical levels used among datasets (not shown).

c. Analysis methods

EOF analysis is used to identify the leading modes of winter AR frequency anomalies over the North Pacific and North Atlantic, with the frequency weighted by the square root of the cosine of latitude. Following Barnes and Polvani (2013) and Ma et al. (2021), the North Pacific and North Atlantic are defined as the regions from 0° to 90°N and 135°E to 125°W and from 0° to 90°N and 60°W to 0°, respectively. Shifting the boundaries of these defined regions in either direction for a few degrees would not affect the results shown in this study. To evaluate the performance of AMIP models on simulating the leading modes in winter AR variability, centered pattern correlation is used to measure the spatial pattern similarity between the modes identified in observation and those derived from AMIP models. As we will show later, AMIP models generally show high skills in reproducing these leading modes, but biases still exist.

To quantify the contributions of SST/sea ice forcing versus internal atmospheric variability in the interannual variability of the observed leading modes, a combined SVD analysis (Wallace et al. 1992; Bretherton et al. 1992) between observation and AMIP ensemble is used in this study. Following Mori et al. (2019), SVD analysis of the covariance matrix of winter AR frequency anomalies between observation and simulations is used to derive the leading modes of interannual AR variability. The spatial patterns derived from this covariance method are quantitatively similar to those from the EOF analysis that is based on the variance of observations or individual simulations. But the leading modes obtained from this covariance method would explain the maximum squared temporal covariances between the observation and simulations over the analysis domains.

To carry out the SVD analysis, we first regrid all detrended winter AR frequency anomalies relative to their own climatology to a common spatial resolution of 1.5° × 1.5°. This can ensure that both observation and simulations have the same spatial dimension. A covariance matrix is then constructed between the observed AR anomalies and the simulated AR anomalies. More specifically, 30 members, one from each of the 30 AMIP models, are concatenated in the time dimension to form a matrix with time dimension size of 1050 (35 winters × 30 members, the row dimension):
M=[model1model30].
To match the time dimension size of 1050 in the simulations, we duplicate the observed time series 30 times to form a matrix for the observed anomalies:
O=[observationobservation].
The product of the two matrices along the time dimension forms a covariance matrix on which we carry out the SVD analysis [Eq. (1)]:
Cov[observation,models]=110501(OMT).
The left and right singular vectors derived from this covariance matrix with largest singular values depict the spatial patterns of the leading modes of interannual winter AR variability in observation and simulations, respectively. The associated expansion coefficients (ECs) are obtained by projecting the detrended winter AR frequency anomalies onto the singular vectors, giving the time series for the 35-winter period in observation (also duplicates 30 times) and simulations in each ensemble member.
As will be shown later (see Fig. 5), we evaluate models’ skill in simulating AR variability, which is defined as the standard deviation of winter AR frequency,
std(f),
where f is the time series of winter AR frequency for a mode with dimension of 35 (i.e., 35 winters). We further decompose the AR variability in models into a forced component driven by boundary forcing,
forcedvariability=std(f¯),
where f¯ is the ensemble-mean time series of winter AR frequency, and an internal variability component,
internalvariability=135[y=135std(fallyf¯y)],
where y denotes the winter of the yth year and all represents all models. For example, fall1 represents the AR frequency of the first winter (i.e., 1980) of all models (i.e., 30).

3. Observed EOF modes and model evaluations

a. The observed leading modes of winter AR variability based on EOF analysis

Figure 1 shows the observed leading EOF modes of winter AR anomalies over the North Pacific and the associated anomalies of sea level pressure, IVT, 850-mb zonal wind and integrated water vapor (IWV). The positive phase of the EOF mode is defined such that the anomalous AR frequency is enhanced over the U.S. west coast. The first mode is characterized by enhanced AR frequency over the AR maximum region and reduced AR activities farther north (Fig. 1a). The spatial pattern of this mode is very similar to the first mode of the North Pacific winter AR variability based on AR IVT identified in Dong et al. (2018; see their Fig. 3c). The band of enhanced AR frequency extends northeastward and depicts higher-than-normal AR frequency along the North American west coast, especially along the U.S west coast. The associated sea level pressure anomaly pattern shows a low anomaly over the northeast Pacific, indicating a southeastward shift of the Aleutian low (Fig. 1c). This low anomaly induces a cyclonic pattern in the IVT anomalies and enhances onshore IVT toward the west coast of North America. In agreement with the presence of a low anomaly over the northeast Pacific, the 850-mb zonal wind strengthens and extends eastward on the equatorward side of the climatological jet and weakens on the poleward side (Fig. 1e). This associated wind anomaly pattern also bears strong resemblance to the first EOF mode of 850-mb zonal wind anomalies over the Pacific (Fig. S1). Indeed, the principal component (PC) associated with AR variability correlates significantly with the PC of the 850-mb zonal wind anomalies with correlation coefficient of 0.78. This implies that the interannual variability in ARs is mostly controlled by the variability in circulation. Moreover, the associated IWV shows positive anomalies over the eastern Pacific near/along the North American west coast (Fig. 1g). Both the positive anomalies in wind and IWV contribute to enhanced AR frequency along the North American west coast. Farther away from the eastern Pacific, negative IWV anomalies are seen nearly everywhere, even over regions with enhanced AR frequency. This again suggests a stronger role wind variability plays in regulating the AR variability.

Fig. 1.
Fig. 1.

(a),(b) Observed AR frequency anomalies, (c),(d) sea level pressure and IVT anomalies, (e),(f) 850-mb zonal wind anomalies, and (g),(h) IWV anomalies associated with the (left) first and (right) second EOF modes of winter AR anomalies over the North Pacific. The values (i.e., 27.9% and 15.4%) in the titles of (a) and (b) indicate the percentage of variance explained by each EOF. Patterns are the regression coefficients obtained by regressing their anomaly fields onto the respective standardized principal components (PCs) (i.e., linear regression between the time series of the anomaly fields and the time series of the standardized PCs). Solid contours in (a)–(f) indicate the climatology of the respective field. Shaded contours show the anomalies. Stippling indicates regions with anomalies significant at the 95% confidence level based on the Student’s t test.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

The second mode depicts a narrow band of slightly enhanced AR frequency poleward of the AR maximum region, accompanied by negative AR anomalies on both sides of this positive AR anomaly (Fig. 1b). This mode also shows strong similarity to the second mode of the North Pacific winter AR variability in Dong et al. (2018; see their Fig. 3f). The associated sea level pressure anomaly pattern shows a high anomaly over the central subtropical Pacific and a low anomaly over the northeast Pacific (Fig. 1d). This dipole pattern in sea level pressure resembles the North Pacific Oscillation (Rogers 1981). Compared to the first mode, the low anomaly is shifted northward and is weaker in intensity (Figs. 1c,d), resulting in weaker IVT anomalies directed toward the northwestern United States and British Columbia. Consistent with the sea level pressure anomalies, the associated wind anomalies show strengthening and eastward extension of the jet (Fig. 1f). Since there is no significant IWV anomaly over the region off the west coast (Fig. 1h), the enhanced AR frequency there is mostly due to the strengthening of the winds.

Since ARs contribute significantly to both the mean and extreme precipitation for many regions around the world, Fig. 2 displays the mean and extreme precipitation patterns associated with these leading AR modes. Consistent with the AR anomaly pattern (Fig. 1a), the first mode is associated with both enhanced mean and extreme precipitation over the eastern Pacific and along the U.S west coast (Figs. 2a,c). Reduced mean and extreme precipitation are found over the central North Pacific from about 30° to 50°N. Significant anomalies are also found in the tropics south of 20°N, where no significant AR anomaly is found. Since most of the tropical precipitation is controlled by convective activities rather than ARs that are typically associated with cold fronts, significant precipitation anomalies over these latitudes likely result from the anomalies of these non-AR weather systems, which covary with the AR mode. As shown in Fig. 2b, the mean precipitation anomalies associated with the second mode show a band of enhanced precipitation over regions at around 40°N stretching across the Pacific from about 180° to the coast of northwestern United States. Significant negative precipitation anomalies are also found over the coastal regions of northwest Pacific and central Pacific equatorward of the positive precipitation band. The negative anomalies over these regions are also significant for the extreme precipitation (Fig. 2d). Note that the GPCP covers a relatively short period. To test the robustness of the results, precipitation from ERA-Interim, which covers the whole period, is used to reproduce the above results. The anomaly patterns based on ERA-Interim are very similar to those based on GPCP, further corroborating the robustness of the results (Fig. S2).

Fig. 2.
Fig. 2.

As in Fig. 1, but for (a),(b) mean precipitation and (c),(d) extreme precipitation anomalies associated with the (left) first and (right) second EOF modes of winter mean AR anomalies over the North Pacific. Precipitation anomaly patterns are obtained by regressing their anomaly fields onto the respective standardized principal components (PCs) (i.e., linear regression between the time series of the precipitation anomaly fields and the time series of the standardized PCs). Stippling indicates regions with anomalies significant at the 95% confidence level based on the Student’s t test.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

We next switch to the leading EOF modes of ARs over the Atlantic, with the positive phase indicating enhanced AR frequency over the western Europe. The first mode shows substantial increases in AR frequency around 30°N stretching from the southeastern United States to the Iberian Peninsula and northwest Africa (Fig. 3a). Reduced AR activities occur over higher latitudes stretching from the northeastern United States to northwestern Europe. Consistent with the first mode of the North Atlantic winter AR variability in Li et al. (2022) (see their Fig. 2a), this pattern indicates the equatorward shift of ARs. Slightly higher-than-normal AR conditions are also found over the Labrador Sea. Moreover, the first AR mode is closely related to the North Atlantic Oscillation (NAO) (Hurrell et al. 2003) or a meridional shift of the Atlantic jet. The associated sea level pressure anomaly pattern shows a negative NAO-like pattern, with a positive anomaly poleward of about 70°N and a low anomaly over the North Atlantic from 30°N to 50°N. This low anomaly induces a cyclonic anomaly in the IVT field, causing strong onshore IVT anomalies over the Iberian Peninsula and northwest Africa (Fig. 3c). Consistent with the equatorward shift of ARs, the associated zonal wind anomaly pattern also shows an equatorward shift of the jet with weakening of the winds on the poleward side and strengthening of the winds on the equatorward side of the jet (Fig. 3e). The associated IWV anomalies tend to have the same signs with the AR anomalies: regions with positive (negative) AR anomalies are also regions with positive (negative) IWV anomalies (Fig. 3g).

Fig. 3.
Fig. 3.

As in Fig. 1, but for the North Atlantic: (a),(b) observed AR frequency anomalies, (c),(d) sea level pressure and IVT anomalies, (e),(f) 850-mb zonal wind anomalies, and (g),(h) IWV anomalies associated with the (left) first and (right) second EOF modes of winter AR anomalies. Patterns are obtained by regressing their anomaly fields onto the respective standardized principal components (PCs). Solid contours in (a)–(f) indicate the climatology. Shaded contours show the anomalies. Stippling indicates regions with anomalies significant at the 95% confidence level based on the Student’s t test.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

As shown in Fig. 3b, the second mode over the Atlantic shows enhanced AR frequency over the AR maximum region and reduced AR frequency on both sides, indicating the narrowing of the AR peak region rather than the shift by the first mode (Fig. 3a). This mode is also identified as the second mode of SST-forced North Atlantic winter AR variability in Li et al. (2022; see their Fig. 4a). Compared to the spatial pattern in Li et al. (2022), the anomalies of this mode identified here extend more northeastward. This mode results in heightened AR activities over most of the coastal regions along western Europe. The associated sea level pressure anomaly pattern shows a dipole structure with a low anomaly located at North Atlantic from 40° to 70°N and a high anomaly from 20° to 40°N. This dipole structure enhances onshore IVT along the Europe west coast and also causes strengthening and eastward extension of the jet (Figs. 3d,f).

The two AR modes over the Atlantic also result in different regional impacts on precipitation. The first mode over the Atlantic is associated with widespread declines in precipitation over the North Atlantic, including large areas along the west coast of Europe and eastern Greenland. Significant decreases in extreme precipitation are only found over the high-latitude regions of North Atlantic at around 70°N. But significant increases in precipitation are found over regions at around 30°N, especially over the Iberian Peninsula (Fig. 4a). The Iberian Peninsula also tends to experience higher-than-normal extreme precipitation (Fig. 4c). Consistent with its AR anomaly pattern (Fig. 3b), the second mode tends to enhance both the mean and extreme precipitation over the Iberian Peninsula and Britain (Figs. 4b,d). Again, the results shown here are consistent with those based on the precipitation data from ERA-Interim (Fig. S3).

Fig. 4.
Fig. 4.

As in Fig. 3, but for (a),(b) mean precipitation and (c),(d) extreme precipitation anomalies associated with the (left) first and (right) second EOF modes of winter AR anomalies over the North Atlantic. Patterns are obtained by regressing their anomaly fields onto the respective standardized principal components (PCs). Stippling indicates regions with anomalies significant at the 95% confidence level based on the Student’s t test.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

b. AMIP model evaluation

Before using the AMIP models to quantify the contributions of SST/sea ice forcing versus internal atmospheric variability to the interannual variability of the leading modes in AR variability, we first evaluate their skills on simulating these modes shown in Figs. 1 and 3. Figures 5a and 5b show the winter AR frequency climatology and AR variability defined as the standard deviation of winter AR frequency in observation and simulations, respectively. Overall, AMIP models can reproduce the climatology of AR distribution very well. They capture both the pattern and magnitude of the AR frequency over both the Atlantic and Pacific, except that the AR frequency over the AR maximum region in the Pacific is slightly underestimated. In terms of interannual variability, AMIP models reproduce both the spatial distribution and magnitude of the variability with high fidelity. They capture the maximum variability on the equatorward side of the climatological AR peak region. Decomposition of the AR variability in models shows that both the boundary-forced variability and the internally driven variability maximize over the equatorward side of the AR peak region (Figs. 5c,d), and that internally driven variability dominates over the boundary-forced variability at nearly all grid points. We also note that, for the leading AR modes, the percentage of variability explained by the boundary forcing can be greater than that for each grid point, which will be elaborated in section 4.

Fig. 5.
Fig. 5.

(a) Observed and (b) simulated ensemble-mean AR climatology (solid contours) and variability (shaded contours). AR variability is defined as the standard deviation of the winter mean AR frequency. (c),(d) The forced and the internally driven components of the total variability in the AMIP models for each grid point. Forced variability is calculated as the standard deviation of the ensemble-mean winter AR frequency across the 35 winters. Internally driven variability is estimated by the multiyear average of the standard deviations across all model anomalies relative to the ensemble mean (i.e., after removing the forced component).

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

We next extract the leading modes of AR variability in the AMIP models using EOF analysis and evaluate how well models simulate the spatial structure of the observed leading modes using centered pattern correlation. Centered pattern correlation is the grid-by-grid spatial correlation between two anomaly spatial maps. The anomalies are computed by removing their own spatial mean from each map (centered). It measures the similarity of the spatial patterns between two maps. When performing EOF analysis on observations and simulations to extract the leading modes of the same field, the order of the leading modes in a model may not match those in observations. To resolve this problem, we apply the EOF swapping method (Lee et al. 2021, 2019). For each ocean basin, we retain the first three leading modes identified in a model and calculate the pattern correlations between the first mode in observation and each of the three leading modes in the model. The mode in a model that has the highest correlation with the observed mode is then identified as the matching mode. Similarly, to identify the corresponding second mode in a model, the pattern correlations between the observed second mode and each of the remaining two modes in the model are calculated. The simulated mode with the highest pattern correlation is identified as the second mode.

Figure 6 shows that AMIP models generally have high skill in reproducing the spatial structure of the observed modes, with the ensemble-mean pattern correlations being 0.85, 0.69, 0.85, and 0.75 for the first and second modes in the Pacific and Atlantic, respectively. Models tend to simulate the first mode of each basin better compared to the second mode. Also, the box-and-whisker plots on the right show that the distribution for the 30-member ensemble, with one member from each of 30 AMIP models, is similar to that of the 40-member ensemble, with 10 members from each of four AMIP models. This suggests that intermember spread within one model is comparable to the intermodel spread, indicating that internal atmospheric variability is an important factor in causing the spread in skills, and that individual model biases play a less important role.

Fig. 6.
Fig. 6.

(left) Centered pattern correlations between observation and simulations of the AR anomalies associated with the (a),(c) first and (b),(d) second EOF modes of AR variability over the North Pacific and North Atlantic, respectively. Values in the lower-left corner indicate the ensemble-mean pattern correlations. Four of the models (ACCESS-ESM, CESM2, IPSL and MIROC6) have 10 members. Their ensemble means are indicated by a red star. (right) Box-and-whisker plots showing the intermodel spreads of the pattern correlations for the 30-member AMIP ensemble with only the first member of each model (“AMIP”) and the 40-member ensemble from the four models that each have 10 members (“AMIP Subset”). The box shows the 25th and 75th percentiles of the spread. The upper and lower whiskers are defined by the formula 75th percentile + 1.5 × (75th percentile − 25th percentile) and 25th percentile −1.5 × (75th percentile − 25th percentile), respectively. Values outside of the whiskers are considered as outliers. The orange horizontal lines show the median while the red stars ensemble means.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

Moreover, the analysis on observation in section 3a has suggested that there is a tight relationship between the AR variability and circulation variability on the interannual time scale. That is, the PCs of the EOF modes derived from AR variability are strongly correlated with the corresponding PCs of the EOF modes derived from 850-mb zonal wind variability. Figure 7 shows that models are able to simulate this tight interannual relationship between AR variability and wind variability. The ensemble-mean correlations for the PCs of the corresponding first and second AR/wind modes over the Pacific and Atlantic are 0.80, 0.69, 0.84, and 0.72, respectively. The observed correlations generally fall within the 25th–75th-percentile range of the intermodel spread, except for the second mode in the Atlantic, which falls at the high end of the model spread.

Fig. 7.
Fig. 7.

As in Fig. 6, but for correlations between the principal components (PCs) associated with the EOF modes of AR anomalies and the corresponding PCs associated with EOF modes of 850-mb zonal wind anomalies for the (a),(c) first and (b),(d) second modes over the North Pacific and North Atlantic, respectively. Values in the lower-left corner indicate the ensemble-mean correlations. Four of the models (ACCESS-ESM, CESM2, IPSL, and MIROC6) have 10 members. Their ensemble means are indicated by a red star. Green dots represent the observed values. (right) Box-and-whisker plots showing the intermodel spreads of the correlations for the 30-member AMIP ensemble (“AMIP”) and the 40-member ensemble from the four models that each have 10 members (“AMIP Subset”). The box shows the 25th and 75th percentile of the spread. The upper and lower whiskers are defined by the formula 75th percentile + 1.5 × (75th percentile −25th percentile) and 25th percentile − 1.5 × (75th percentile − 25th percentile), respectively. Values outside of the whiskers are considered as outliers. The orange horizontal lines show the median while the red stars indicate ensemble means.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

In summary, the above results on the model evaluation suggest that AMIP models are capable of simulating the observed winter AR variability over the Northern Hemisphere with high skill, justifying their uses to further investigate the roles of SST/sea ice forcing versus internal atmospheric variability in driving the observed AR variability.

4. Internal atmospheric variability versus SST/sea ice forced variability

a. SVD modes of ARs and associated SST patterns

Having evaluated the EOF modes of individual models with the observed EOF modes, we next extract the SVD modes of ARs based on the interannual covariance between observation and models. This SVD analysis yields a pair of spatial patterns with the largest squared temporal covariances between observation and models. The SVD patterns for observation with the largest squared covariances resemble the observed EOF patterns that explain the largest percentages of AR variance (Figs. 1, 3, and 8), and the SVD patterns for models provide estimates for all the AMIP models.

Fig. 8.
Fig. 8.

Leading modes of AR anomalies obtained from the SVD analysis of the covariance between observation and simulations: the (a),(b) first and (c),(d) second modes of AR anomalies over the North Pacific, and the (e),(f) first and (g),(h) second modes of AR anomalies over the North Atlantic, where the left and right panels display results from (left) observation and (right) simulations. The values in the title of each subpanel indicate the percentage of squared covariance explained by each SVD. Anomaly patterns in observation are obtained by regressing AR anomalies onto the standardized expansion coefficients (ECs) of 35 winters. Regressions in simulations are based on the 1050 winters across all ensemble members (30 members; each has 35 winters). Solid contours indicate climatology of winter AR frequency. Shaded contours represent anomalies. Stippling indicates regions with anomalies significant at the 95% confidence level based on the Student’s t test.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

Figure 8 shows that the SVD patterns of the AMIP models successfully reproduce the spatial structures in observation over both the Atlantic and Pacific, although there are minor differences between observation and simulations for the first mode over the Pacific and the second mode over the Atlantic. Specifically, for the first mode in the Pacific, models are able to capture the negative anomalies over the central North Pacific north of 30°N. They also produce a band of strong positive anomalies at the AR peak region. However, the positive anomalies in the ERA-Interim are tilted more northeastward and impinging toward the northwestern United States, while the positive anomalies in the AMIP models are more zonally oriented, making California the region mostly affected by this mode. For the second mode over the Atlantic, despite models being able to reproduce the tripole pattern with positive anomalies at the AR peak region and negative anomalies on both sides, the positive anomalies in the models shift southwestward compared to observation.

The SST anomaly patterns corresponding to these leading SVD modes are displayed in Fig. 9 by regressing the SST anomalies onto the standardized ECs of the observation and the ensemble-mean EC of the AMIP models for the 35 winters. Over the Pacific, the first mode of AR variability is associated with an SST pattern that resembles the positive phase of ENSO (Figs. 9a,b). Indeed, the observed Niño-3.4 index is strongly correlated with the ensemble-mean EC of the AMIP models at 0.76. Widespread warm anomalies extend from the eastern equatorial Pacific and reach the western equatorial Pacific. The warm anomalies spread toward the extratropics along the eastern Pacific and reach the North Pacific above 40°N. Encircled by the positive anomalies, negative SST anomalies originating from western equatorial Pacific extend northeastward and southeastward in the Northern and Southern Hemispheres, respectively. Significant positive SST anomalies are also found in the Indian Ocean. Moreover, the SST anomalies associated with the second AR mode also show positive anomalies over the equatorial eastern Pacific (Figs. 9c,d). However, compared to the pattern associated with the first mode, the positive anomalies are more confined zonally in observation (Fig. 9c), and both zonally and meridionally in models (Fig. 9d). The negative anomalies originating from the equatorial western Pacific extend more eastward and northward. This SST anomaly pattern over the Pacific bears strong resemblance to the North Pacific mode (NPM) (Deser and Blackmon 1995; Park et al. 2012). Widespread positive anomalies are also found over the Indian Ocean and western Pacific along the East Asian coast. Further analysis shows that the SST patterns over the Pacific associated with the first and second modes of AR variability also resemble the first and second leading EOF modes of the observed interannual SST variability over the North Pacific (Fig. S4).

Fig. 9.
Fig. 9.

As in Fig. 8, but for SST anomalies associated with the leading modes of AR anomalies obtained from the SVD analysis of the covariance between observation and simulations: the (a),(b) first and (c),(d) second modes of SST anomalies over the North Pacific, and the (e),(f) first and (g),(h) second modes of SST anomalies over the North Atlantic. Columns show results from (left) observation and (right) simulations. Anomaly patterns in observation are obtained by regressing the SST anomalies onto the expansion coefficients (ECs) of the 35 winters. Anomalies in simulations are based on the ensemble-mean EC. Stippling indicates regions with anomalies significant at the 95% confidence level based on the Student’s t test.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

For the Atlantic AR modes (Figs. 9e,f,g,h), the associated SST anomaly patterns over the Pacific are very similar to those associated with the AR modes of the Pacific, but with slightly smaller magnitude and reduced areas with significant anomalies. Stronger and significant SST anomalies are found over the Atlantic, highlighting an important role that Atlantic SST plays in modulating the AR variability locally.

The results in Fig. 9 also indicate that the SST patterns in observation tend to be less coherent and significant compared to the SST patterns obtained from simulations, and the differences are especially pronounced for the Atlantic modes (Figs. 9e,g). Since the SST patterns from simulations are obtained by regressing the SST anomalies onto the ensemble-mean EC, these SST patterns are thus associated mostly with the forced AR variability. Therefore, the differences between SST patterns in observation and simulations imply a large fraction of internally driven variability contained in the observed ECs, especially for the modes in the Atlantic.

b. Relative roles of SST/sea ice forcing versus internal atmospheric variability in driving AR variability

We next quantify the roles of SST/sea ice forcing versus internal atmospheric variability in the leading modes of AR interannual variability using the coefficient of the determination (r2) approach (Fig. 10). Indeed, internal atmospheric variability plays an important role in driving the AR variability over both the Pacific and Atlantic as indicated by the large intermodel spread in the ECs. Over the Pacific, about half of the variance in the leading modes can be explained by the SST forced variability, but that fraction drops to about a quarter for the Atlantic modes, confirming that internal atmospheric variability indeed plays a more important role in driving the AR variability over the Atlantic.

Fig. 10.
Fig. 10.

Expansion coefficients (ECs) for the (a),(c) first and (b),(d) second SVD modes of AR anomalies over the (a),(b) North Pacific and (c),(d) North Atlantic. Black and teal curves show ECs for the observation and ensemble-mean EC for simulations, respectively. The R2 values between the observed ECs and the corresponding simulated ensemble-mean EC are shown at the lower-left corners of each panel. The shading shows the 2.5th–97.5th-percentile range of the model spread.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

Given the important role that internal atmospheric variability plays in the AR variability, can it alone drive the formation of the leading AR modes shown in Fig. 8? To answer this question, we apply EOF analysis to the intra-ensemble AR anomalies based on model simulations. To obtain the intra-ensemble AR anomalies, for each model we first removed its own winter AR frequency climatology from the winter AR frequency time series. After that, the ensemble-mean anomalous winter AR frequency was calculated, which represents the forced AR variability. This forced component was then removed from the anomalous winter AR frequency of each model to get the internally driven component. Finally, we concatenated the internally driven components from all models together to obtain the intra-ensemble AR anomalies. Results from this analysis identify leading modes nearly identical to those based on the SVD analysis of the covariance between observation and models or the EOF of individual models or observation (cf. Figs. 1, 3, and 8 and Fig. S5), suggesting that the leading modes forced by SST can also occur naturally due to internal atmospheric variability. These results are also consistent with Ma et al. (2021) in which they discover similar leading modes of AR variability by using an ensemble of models driven by the climatological annual cycle of SST from the Polar Amplification Model Intercomparison Project (PAMIP).

Given that both SST/sea ice forcing and internal atmospheric variability play such important roles in shaping the AR variability, their contributions to the magnitude of the total variance are further quantified. We estimate the total variance by calculating the variance of the ECs corresponding to the leading SVD modes during the 35 winters. The total variance of each mode in models is based on ECs across all models. Figure 11 indicates that AMIP models underestimate the variance for all modes, with the largest underestimate for the second Atlantic mode. We further estimate the magnitude of the forced variance in models by calculating the variance of the ensemble-mean EC of each mode. The forced variance accounts for about half of the total variance for the Pacific modes, while that number drops to about 40% and 30% for the first and second Atlantic modes, respectively. These results further support that AR variability over the Atlantic is more strongly modulated by internal atmospheric variability compared to the Pacific AR variability.

Fig. 11.
Fig. 11.

Magnitude of the total variance (yellow bars) in observation and simulations (blue bars). Red stars indicate the magnitude of the forced variance. Markers indicate the contributions of different SST modes to the total variance of the observed leading SVD modes of AR anomalies. The values for each mode are scaled by the observed total variance. See the text for more information on how the statistics shown in this figure are calculated.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

Last, we investigate how much of the total observed variability in the AR EOF modes is regulated by the leading EOF modes of SST over each of the ocean basins by using the coefficient of determination (r2) approach. The first SST mode over the Pacific associated with ENSO significantly correlates with both the first modes of the Pacific and Atlantic and explains about 30% and 17% of their observed total variance, respectively. The second SST mode over the Pacific, which resembles the NPM, exerts strong influences on the second Pacific AR mode, and explains about half of the observed total variance of this mode. Moreover, we perform EOF analysis on the SST anomalies over the North Atlantic (Fig. S6). The first mode shows a tripole pattern of SST anomalies over the North Atlantic, which resembles the SST anomaly pattern of the second AR mode over the Atlantic. Similar to the SST anomaly pattern of the first AR mode over the North Atlantic, the second mode in SST also shows a dipole pattern over the North Atlantic. Consequently, the leading modes of AR variability over the North Atlantic are also related to the leading SST modes over the North Atlantic. The first and second SST modes in the Atlantic explains about 20% of the observed total variance of the second and first modes of AR variability, respectively.

c. Tropical versus extratropical SST in driving the forced AR variability

Based on the SST patterns associated with the leading modes of AR variability (Fig. 9), significant SST anomalies can be seen over both the tropical and extratropical oceans, suggesting that both play roles in driving the forced AR variability. To better understand the relative roles of tropical SST versus extratropical SST in driving the forced AR variability, we analyze two additional ensembles based on CESM1. One ensemble with 10 members, referred to as GOGA, is driven by observed SST globally. Another ensemble also with 10 members, referred to as TOGA, is driven by observed SST only in the tropics. Despite their smaller ensemble size compared to the 30-member AMIP ensemble, they are still useful in providing an estimate on the relative roles of tropical SST versus extratropical SST in AR variability. The same combined SVD analysis is applied to the covariance matrix between observation and simulations based on GOGA.

Results show that GOGA is able to reproduce both the leading SVD modes of AR variability over the Pacific and the first mode over the Atlantic. The strengthening and eastward extension of ARs associated with the second mode over the Atlantic are too weak in the models (Fig. S7). Compared to the results based on the AMIP ensemble, the 10-member GOGA underestimates the fraction of forced variance in the first Pacific mode while overestimating that in the second Atlantic mode (Fig. 12). The fraction of forced variance for the second Pacific mode and first Atlantic mode are consistent with those based on the AMIP ensemble. They explain about half and a quarter of the total variance, respectively, for the second Pacific mode and the first Atlantic mode. Consistent with the results based on the AMIP ensemble, AR variability in the Pacific is more strongly controlled by the SST variability compared to the AR variability in the Atlantic. For the forced variability, the role of the tropical SST variability dominates over the extratropical SST variability for both the leading modes over the Pacific, explaining about 77% and 71% of the total forced variability, respectively. Over the Atlantic, tropical SST explains about 49% and 56% of the total forced variability for the first and second mode, respectively, suggesting comparable role tropical SST versus extratropical SST plays in driving these modes.

Fig. 12.
Fig. 12.

As in Fig. 10, but for the results based on the GOGA and TOGA ensembles.

Citation: Journal of Climate 35, 23; 10.1175/JCLI-D-22-0089.1

5. Conclusions

A better understanding of the mechanisms controlling the interannual AR variability is crucial for water resource management and hazard warning (Paltan et al. 2017; Henn et al. 2020; Dettinger et al. 2011; Ralph et al. 2006). In this study, we identify and characterize the leading modes of AR variability over both the Pacific and Atlantic during boreal winter, with distinct precipitation patterns and impacts over different coastal regions, similar to Dong et al. (2018) and Li et al. (2022). Over the Pacific, the first mode is characterized by a dipole structure with substantial enhanced AR frequency over the AR peak region and reduced AR activities farther north (Fig. 1a). The second mode depicts a tripole pattern with a band of positive anomalies extending from the west Pacific at about 30°N to the northeast Pacific, accompanied by negative anomalies on both sides (Fig. 1b). The first AR mode in the Pacific significantly modulates precipitation along the U.S. west coast while the second mode mostly affects precipitation over the U.S. Northwest and British Columbia (Fig. 2). Over the Atlantic, the first mode represents the equatorward shift of the ARs, with significant impacts along the entire west coast of Europe, while the second mode represents strengthening and eastward extension of the AR peak region, with the effects confined more toward the southern part of the west coast (Figs. 3a,b). It is worth mentioning that, given the strong correlation between the leading modes of AR variability and the corresponding leading modes of 850-mb zonal wind variability, the spatial patterns of the leading AR modes identified in this study do not depend on the AR detection algorithms used. Indeed, these leading modes of AR variability can also be identified in nearly all of the tier 1 AR catalogues based on the global AR detection algorithms participated in the Atmospheric River Tracking Method Intercomparison Project (not shown; Shields et al. 2018). As such, the AR modes identified here can serve as metrics to evaluate the AR variability in climate models.

We have quantified, perhaps for the first time, the relative roles of SST forcing versus internal atmospheric variability in driving the interannual variability of the leading AR modes by using a large ensemble of AMIP-type models that participated in CMIP6. We first show that the AMIP models have high skill in simulating many aspects of these modes, including their spatial structure and the strong coupling between circulation variability and AR variability.

Then, a combined SVD analysis using both observation and AMIP models is adopted to extract the leading modes of AR covariability between observation and models. Based on the SVD analysis, the forced variability explains about half of and a quarter of the total variance for the leading AR modes over the Pacific and the Atlantic, respectively. These results suggest higher predictability of the AR variability over the Pacific compared to the Atlantic. The associated SST patterns for the leading modes over the Pacific bear marked resemblance to the corresponding leading modes of North Pacific SST, characterized by an ENSO-like (Xiong and Ren 2021; Kim et al. 2019) and an NPM-like (Deser and Blackmon 1995; Park et al. 2012) anomalous SST pattern, respectively. Further analysis reveals that these two Pacific SST modes also play some role in the corresponding leading modes of Atlantic AR variability. Additionally, a model ensemble driven by observed SST only in the tropical oceans is used to quantify the relative importance of tropical versus extratropical SST in exciting these leading modes of AR variability. Results show that tropical SST variability dominates over extratropical SST variability and contributes more to the total forced AR variability over the Pacific. Over the Atlantic, the tropical and extratropical SST play comparable roles in the formations of the leading modes of AR variability.

Our findings may have important implications for the interannual variations of boreal winter precipitation. While the SST variability modes, such as ENSO or NPM, provide some sources of predictability for the North American west coast, the predictability is limited by the internal variability of the atmosphere (Teng and Branstator 2017; Zhang et al. 2018; Dong et al. 2018; Kumar and Chen 2017; Chen and Kumar 2018; Deser et al. 2018; Cash and Burls 2019), especially for the well-known failed 2016 winter rains in Southern California (Zhang et al. 2018; Patricola et al. 2020). Dong et al. (2018) estimate that about 80% of the interannual variability of winter precipitation in California is controlled by internal atmospheric variability, but we find that internal atmospheric variability only accounts for about half of the variability in the two Pacific AR modes. As both AR modes can influence the North American west coast (Fig. 2), this indicates that the winter precipitation related to the two AR modes may be more predictable. Our analysis also indicates that SST variability over the Pacific basin can influence the AR variability not only locally over the Pacific, but also remotely over the Atlantic (Fig. 9), similar to the remote influence of ENSO on NAO (Li and Lau 2012; Brönnimann 2007; Zhang et al. 2019). Moreover, the leading AR modes forced by SST can appear as the dominant pattern in the AR response to Arctic sea ice loss (Ma et al. 2021) and also occur naturally due to internal atmospheric variability (Fig. S5). Further research is needed to better understand the mechanisms of internal atmospheric variability in driving AR variability at shorter time scales. This can potentially improve AR forecast at S2S time scale. Overall, the results presented in this study not only improve our physical understanding of the AR variability, but may also lead to better seasonal AR prediction over the densely populated regions, such as the North American west coast and western Europe.

Acknowledgments.

Weiming Ma and Gang Chen are supported by National Science Foundation Grants AGS-1832842 and NASA Grant 80NSS-C21K1522. We thank three anonymous reviewers for their constructive feedback that greatly improved the manuscript. The authors would like to acknowledge high-performance computing support from Cheyenne (https://doi.org/10.5065/D6RX99HX) provided by NCAR’s Computational and Information Systems Laboratory, sponsored by the National Science Foundation.

Data availability statement.

Data from the AMIP models participating in CMIP6 can be found on the Earth System Grid at https://esgf-node.llnl.gov/projects/cmip6/. Data from GOGA can be found at https://www.earthsystemgrid.org/dataset/ucar.cgd.ccsm4.cam5.prescribed_sst_amip.goga.html. Data from TOGA can be found at https://www.earthsystemgrid.org/dataset/ucar.cgd.ccsm4.cam5.prescribed_sst_amip.toga.html. ERA-Interim is available for download through https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era-interim/. The SST dataset can be found at https://www.metoffice.gov.uk/hadobs/hadisst/. The GPCP dataset can be found at https://www.ncei.noaa.gov/products/climate-data-records/precipitation-gpcp-daily.

REFERENCES

  • Baggett, C. F., E. A. Barnes, E. D. Maloney, and B. D. Mundhenk, 2017: Advancing atmospheric river forecasts into subseasonal-to-seasonal time scales. Geophys. Res. Lett., 44, 75287536, https://doi.org/10.1002/2017GL074434.

    • Search Google Scholar
    • Export Citation
  • Barnes, E. A., and L. Polvani, 2013: Response of the midlatitude jets, and of their variability, to increased greenhouse gases in the CMIP5 models. J. Climate, 26, 71177135, https://doi.org/10.1175/JCLI-D-12-00536.1.

    • Search Google Scholar
    • Export Citation
  • Bond, N. A., M. F. Cronin, H. Freeland, and N. Mantua, 2015: Causes and impacts of the 2014 warm anomaly in the NE Pacific. Geophys. Res. Lett., 42, 34143420, https://doi.org/10.1002/2015GL063306.

    • Search Google Scholar
    • Export Citation
  • Bozkurt, D., O. L. Sen, Y. Ezber, B. Guan, M. Viale, and F. Caglar, 2021: Influence of African atmospheric rivers on precipitation and snowmelt in the Near East’s highlands. J. Geophys. Res. Atmos., 126, e2020JD033646, https://doi.org/10.1029/2020JD033646.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., C. Smith, and J. M. Wallace, 1992: An intercomparison of methods for finding coupled patterns in climate data. J. Climate, 5, 541560, https://doi.org/10.1175/1520-0442(1992)005<0541:AIOMFF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brönnimann, S., 2007: Impact of El Niño–Southern Oscillation on European climate. Rev. Geophys., 45, RG3003, https://doi.org/10.1029/2006RG000199.

    • Search Google Scholar
    • Export Citation
  • Cash, B. A., and N. J. Burls, 2019: Predictable and unpredictable aspects of U.S. West Coast rainfall and El Niño: Understanding the 2015/16 event. J. Climate, 32, 28432868, https://doi.org/10.1175/JCLI-D-18-0181.1.

    • Search Google Scholar
    • Export Citation
  • Chen, M., and A. Kumar, 2018: Winter 2015/16 atmospheric and precipitation anomalies over North America: El Niño response and the role of noise. Mon. Wea. Rev., 146, 909927, https://doi.org/10.1175/MWR-D-17-0116.1.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, https://doi.org/10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Deser, C., and M. Blackmon, 1995: On the relationship between tropical and North Pacific sea surface temperature variations. J. Climate, 8, 16771680, https://doi.org/10.1175/1520-0442(1995)008<1677:OTRBTA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Deser, C., I. R. Simpson, A. S. Phillips, and K. A. McKinnon, 2018: How well do we know ENSO’s climate impacts over North America, and how do we evaluate models accordingly? J. Climate, 31, 49915014, https://doi.org/10.1175/JCLI-D-17-0783.1.

    • Search Google Scholar
    • Export Citation
  • Dettinger, M. D., F. M. Ralph, T. Das, P. J. Neiman, and D. R. Cayan, 2011: Atmospheric rivers, floods and the water resources of California. Water, 3, 445478, https://doi.org/10.3390/w3020445.

    • Search Google Scholar
    • Export Citation
  • Dong, L., L. R. Leung, F. Song, and J. Lu, 2018: Roles of SST versus internal atmospheric variability in winter extreme precipitation variability along the U.S. West Coast. J. Climate, 31, 80398058, https://doi.org/10.1175/JCLI-D-18-0062.1.

    • Search Google Scholar
    • Export Citation
  • Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 19371958, https://doi.org/10.5194/gmd-9-1937-2016.

    • Search Google Scholar
    • Export Citation
  • Francis, D., K. S. Mattingly, M. Temimi, R. Massom, and P. Heil, 2020: On the crucial role of atmospheric rivers in the two major Weddell Polynya events in 1973 and 2017 in Antarctica. Sci. Adv., 6, eabc2695, https://doi.org/10.1126/sciadv.abc2695.

    • Search Google Scholar
    • Export Citation
  • Gao, Y., J. Lu, and L. R. Leung, 2016: Uncertainties in projecting future changes in atmospheric rivers and their impacts on heavy precipitation over Europe. J. Climate, 29, 67116726, https://doi.org/10.1175/JCLI-D-16-0088.1.

    • Search Google Scholar
    • Export Citation
  • Gershunov, A., T. Shulgina, F. M. Ralph, D. A. Lavers, and J. J. Rutz, 2017: Assessing the climate-scale variability of atmospheric rivers affecting western North America. Geophys. Res. Lett., 44, 79007908, https://doi.org/10.1002/2017GL074175.

    • Search Google Scholar
    • Export Citation
  • Gorodetskaya, I. V., M. Tsukernik, K. Claes, M. F. Ralph, W. D. Neff, and N. P. M. Van Lipzig, 2014: The role of atmospheric rivers in anomalous snow accumulation in East Antarctica. Geophys. Res. Lett., 41, 61996206, https://doi.org/10.1002/2014GL060881.

    • Search Google Scholar
    • Export Citation
  • Guan, B., and D. E. Waliser, 2015: Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. J. Geophys. Res. Atmos., 120, 12 51412 535, https://doi.org/10.1002/2015JD024257.

    • Search Google Scholar
    • Export Citation
  • Guan, B., D. E. Waliser, N. P. Molotch, E. J. Fetzer, and P. J. Neiman, 2012: Does the Madden–Julian oscillation influence wintertime atmospheric rivers and snowpack in the Sierra Nevada? Mon. Wea. Rev., 140, 325342, https://doi.org/10.1175/MWR-D-11-00087.1.

    • Search Google Scholar
    • Export Citation
  • Guan, B., D. E. Waliser, F. M. Ralph, E. J. Fetzer, and P. J. Neiman, 2016: Hydrometeorological characteristics of rain-on-snow events associated with atmospheric rivers. Geophys. Res. Lett., 43, 29642973, https://doi.org/10.1002/2016GL067978.

    • Search Google Scholar
    • Export Citation
  • Guirguis, K., A. Gershunov, T. Shulgina, R. E. S. Clemesha, and F. M. Ralph, 2019: Atmospheric rivers impacting Northern California and their modulation by a variable climate. Climate Dyn., 52, 65696583, https://doi.org/10.1007/s00382-018-4532-5.

    • Search Google Scholar
    • Export Citation
  • Henn, B., K. N. Musselman, L. Lestak, F. M. Ralph, and N. P. Molotch, 2020: Extreme runoff generation from atmospheric river driven snowmelt during the 2017 Oroville Dam spillways incident. Geophys. Res. Lett., 47, e2020GL088189, https://doi.org/10.1029/2020GL088189.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., R. F. Adler, M. M. Morrissey, D. T. Bolvin, S. Curtis, R. Joyce, B. McGavock, and J. Susskind, 2001: Global precipitation at one-degree daily resolution from multisatellite observations. J. Hydrometeor., 2, 3650, https://doi.org/10.1175/1525-7541(2001)002<0036:GPAODD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., Y. Kushnir, G. Ottersen, and M. Visbeck, 2003: An overview of the North Atlantic Oscillation. Climatic Significance and Environmental Impact, Geophys. Monogr., No. 134, Amer. Geophys. Union, 135.

  • Kim, H.-M., Y. Zhou, and M. A. Alexander, 2019: Changes in atmospheric rivers and moisture transport over the Northeast Pacific and western North America in response to ENSO diversity. Climate Dyn., 52, 7375–7388, https://doi.org/10.1007/s00382-017-3598-9.

    • Search Google Scholar
    • Export Citation
  • Kumar, A., and M. Chen, 2017: What is the variability in US west coast winter precipitation during strong El Niño events? Climate Dyn., 49, 27892802, https://doi.org/10.1007/s00382-016-3485-9.

    • Search Google Scholar
    • Export Citation
  • Lamjiri, M. A., M. D. Dettinger, F. M. Ralph, and B. Guan, 2017: Hourly storm characteristics along the U.S. West Coast: Role of atmospheric rivers in extreme precipitation. Geophys. Res. Lett., 44, 70207028, https://doi.org/10.1002/2017GL074193.

    • Search Google Scholar
    • Export Citation
  • Lavers, D. A., and G. Villarini, 2013: The nexus between atmospheric rivers and extreme precipitation across Europe. Geophys. Res. Lett., 40, 32593264, https://doi.org/10.1002/grl.50636.

    • Search Google Scholar
    • Export Citation
  • Lavers, D. A., and G. Villarini, 2015: The contribution of atmospheric rivers to precipitation in Europe and the United States. J. Hydrol., 522, 382390, https://doi.org/10.1016/j.jhydrol.2014.12.010.

    • Search Google Scholar
    • Export Citation
  • Lee, J., K. R. Sperber, P. J. Gleckler, C. J. W. Bonfils, and K. E. Taylor, 2019: Quantifying the agreement between observed and simulated extratropical modes of interannual variability. Climate Dyn., 52, 40574089, https://doi.org/10.1007/s00382-018-4355-4.

  • Lee, J., K. R. Sperber, P. J. Gleckler, K. E. Taylor, and C. J. W. Bonfils, 2021: Benchmarking performance changes in the simulation of extratropical modes of variability across CMIP generations. J. Climate, 34, 6945–6969, https://doi.org/10.1175/JCLI-D-20-0832.1.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., and Y. Qian, 2009: Atmospheric rivers induced heavy precipitation and flooding in the western U.S. simulated by the WRF regional climate model. Geophys. Res. Lett., 36, L03820, https://doi.org/10.1029/2008GL036445.

  • Li, C., W. Mei, and Y. Kamae, 2022: Variability and predictability of cold-season North Atlantic atmospheric river occurrence frequency in a set of high-resolution atmospheric simulations. Climate Dyn., 58, 24852500, https://doi.org/10.1007/s00382-021-06017-y.

    • Search Google Scholar
    • Export Citation
  • Li, Y., and N.-C. Lau, 2012: Impact of ENSO on the atmospheric variability over the North Atlantic in late winter—Role of transient eddies. J. Climate, 25, 320342, https://doi.org/10.1175/JCLI-D-11-00037.1.

    • Search Google Scholar
    • Export Citation
  • Ma, W., G. Chen, and B. Guan, 2020a: Poleward shift of atmospheric rivers in the Southern Hemisphere in recent decades. Geophys. Res. Lett., 47, e2020GL089934, https://doi.org/10.1029/2020GL089934.

    • Search Google Scholar
    • Export Citation
  • Ma, W., J. Norris, and G. Chen, 2020b: Projected changes to extreme precipitation along North American West Coast from the CESM large ensemble. Geophys. Res. Lett., 47, e2019GL086038, https://doi.org/10.1029/2019GL086038.

    • Search Google Scholar
    • Export Citation
  • Ma, W., G. Chen, Y. Peings, and N. Alviz, 2021: Atmospheric river response to Arctic sea ice loss in the Polar Amplification Model Intercomparison Project. Geophys. Res. Lett., 48, e2021GL094883, https://doi.org/10.1029/2021GL094883.

    • Search Google Scholar
    • Export Citation
  • Mattingly, K. S., T. L. Mote, and X. Fettweis, 2018: Atmospheric river impacts on Greenland Ice Sheet surface mass balance. J. Geophys. Res. Atmos., 123, 85388560, https://doi.org/10.1029/2018JD028714.

    • Search Google Scholar
    • Export Citation
  • McGowan, H., K. Borthwick, A. Schwartz, J. N. Callow, S. Bilish, and S. Browning, 2021: Atmospheric rivers: An overlooked threat to the marginal snowpack of the Australian Alps. J. Hydrometeor., 22, 25212532, https://doi.org/10.1175/JHM-D-20-0293.1.

    • Search Google Scholar
    • Export Citation
  • Mori, M., Y. Kosaka, M. Watanabe, H. Nakamura, and M. Kimoto, 2019: A reconciled estimate of the influence of Arctic sea-ice loss on recent Eurasian cooling. Nat. Climate Change, 9, 123129, https://doi.org/10.1038/s41558-018-0379-3.

    • Search Google Scholar
    • Export Citation
  • Mundhenk, B. D., E. A. Barnes, E. D. Maloney, and C. F. Baggett, 2018: Skillful empirical subseasonal prediction of landfalling atmospheric river activity using the Madden–Julian oscillation and quasi-biennial oscillation . npj Climate Atmos. Sci., 1, 20177, https://doi.org/10.1038/s41612-017-0008-2.

    • Search Google Scholar
    • Export Citation
  • Nash, D., D. Waliser, B. Guan, H. Ye, and F. M. Ralph, 2018: The role of atmospheric rivers in extratropical and polar hydroclimate. J. Geophys. Res. Atmos., 123, 68046821, https://doi.org/10.1029/2017JD028130.

    • Search Google Scholar
    • Export Citation
  • Paltan, H., D. Waliser, W. H. Lim, B. Guan, D. Yamazaki, R. Pant, and S. Dadson, 2017: Global floods and water availability driven by atmospheric rivers. Geophys. Res. Lett., 44, 10 38710 395, https://doi.org/10.1002/2017GL074882.

    • Search Google Scholar
    • Export Citation
  • Park, J.-Y., S.-W. Yeh, and J.-S. Kug, 2012: Revisited relationship between tropical and North Pacific sea surface temperature variations. Geophys. Res. Lett., 39, L02703, https://doi.org/10.1029/2011GL050005.

    • Search Google Scholar
    • Export Citation
  • Patricola, C. M., J. P. O’Brien, M. D. Risser, A. M. Rhoades, T. A. O’Brien, P. A. Ullrich, D. A. Stone, and W. D. Collins, 2020: Maximizing ENSO as a source of western US hydroclimate predictability. Climate Dyn., 54, 351372, https://doi.org/10.1007/s00382-019-05004-8.

    • Search Google Scholar
    • Export Citation
  • Ralph, F. M., P. J. Neiman, G. A. Wick, S. I. Gutman, M. D. Dettinger, D. R. Cayan, and A. B. White, 2006: Flooding on California’s Russian River: Role of atmospheric rivers. Geophys. Res. Lett., 33, L13801, https://doi.org/10.1029/2006GL026689.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., 1981: The North Pacific Oscillation. J. Climatol., 1, 3957, https://doi.org/10.1002/joc.3370010106.

  • Rutz, J. J., and W. J. Steenburgh, 2012: Quantifying the role of atmospheric rivers in the interior western United States. Atmos. Sci. Lett., 13, 257261, https://doi.org/10.1002/asl.392.

    • Search Google Scholar
    • Export Citation
  • Shields, C. A., and Coauthors, 2018: Atmospheric River Tracking Method Intercomparison Project (ARTMIP): Project goals and experimental design. Geosci. Model Dev., 11, 24552474, https://doi.org/10.5194/gmd-11-2455-2018.

    • Search Google Scholar
    • Export Citation
  • Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485498, https://doi.org/10.1175/BAMS-D-11-00094.1.

    • Search Google Scholar
    • Export Citation
  • Teng, H., and G. Branstator, 2017: Causes of extreme ridges that induce California droughts. J. Climate, 30, 14771492, https://doi.org/10.1175/JCLI-D-16-0524.1.

    • Search Google Scholar
    • Export Citation
  • Thapa, K., T. A. Endreny, and C. R. Ferguson, 2018: Atmospheric rivers carry nonmonsoon extreme precipitation into Nepal. J. Geophys. Res. Atmos., 123, 59015912, https://doi.org/10.1029/2017JD027626.

    • Search Google Scholar
    • Export Citation
  • Viale, M., R. Valenzuela, R. D. Garreaud, and F. M. Ralph, 2018: Impacts of atmospheric rivers on precipitation in southern South America. J. Hydrometeor., 19, 16711687, https://doi.org/10.1175/JHM-D-18-0006.1.

    • Search Google Scholar
    • Export Citation
  • Waliser, D., and B. Guan, 2017: Extreme winds and precipitation during landfall of atmospheric rivers. Nat. Geosci., 10, 179183, https://doi.org/10.1038/ngeo2894.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., C. Smith, and C. S. Bretherton, 1992: Singular value decomposition of wintertime sea surface temperature and 500-mb height anomalies. J. Climate, 5, 561576, https://doi.org/10.1175/1520-0442(1992)005<0561:SVDOWS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Woods, C., and R. Caballero, 2016: The role of moist intrusions in winter Arctic warming and sea ice decline. J. Climate, 29, 44734485, https://doi.org/10.1175/JCLI-D-15-0773.1.

    • Search Google Scholar
    • Export Citation
  • Xiong, Y., and X. Ren, 2021: Influences of atmospheric rivers on North Pacific winter precipitation: Climatology and dependence on ENSO condition. J. Climate, 34, 277292, https://doi.org/10.1175/JCLI-D-20-0301.1.

    • Search Google Scholar
    • Export Citation
  • Zhang, P., G. Chen, W. Ma, Y. Ming, and Z. Wu, 2021: Robust atmospheric river response to global warming in idealized and comprehensive climate models. J. Climate, 34, 77177734, https://doi.org/10.1175/JCLI-D-20-1005.1.

    • Search Google Scholar
    • Export Citation
  • Zhang, T., and Coauthors, 2018: Predictability and prediction of Southern California rains during strong El Niño events: A focus on the failed 2016 winter rains. J. Climate, 31, 555574, https://doi.org/10.1175/JCLI-D-17-0396.1.

    • Search Google Scholar
    • Export Citation
  • Zhang, W., Z. Wang, M. F. Stuecker, A. G. Turner, F.-F. Jin, and X. Geng, 2019: Impact of ENSO longitudinal position on teleconnections to the NAO. Climate Dyn., 52, 257274, https://doi.org/10.1007/s00382-018-4135-1.

    • Search Google Scholar
    • Export Citation
  • Zhou, Y., H. Kim, and D. E. Waliser, 2021: Atmospheric river lifecycle responses to the Madden–Julian oscillation. Geophys. Res. Lett., 48, e2020GL090983, https://doi.org/10.1029/2020GL090983.

    • Search Google Scholar
    • Export Citation
  • Zhu, Y., and R. E. Newell, 1998: A proposed algorithm for moisture fluxes from atmospheric rivers. Mon. Wea. Rev., 126, 725735, https://doi.org/10.1175/1520-0493(1998)126<0725:APAFMF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

Supplementary Materials

Save
  • Baggett, C. F., E. A. Barnes, E. D. Maloney, and B. D. Mundhenk, 2017: Advancing atmospheric river forecasts into subseasonal-to-seasonal time scales. Geophys. Res. Lett., 44, 75287536, https://doi.org/10.1002/2017GL074434.

    • Search Google Scholar
    • Export Citation
  • Barnes, E. A., and L. Polvani, 2013: Response of the midlatitude jets, and of their variability, to increased greenhouse gases in the CMIP5 models. J. Climate, 26, 71177135, https://doi.org/10.1175/JCLI-D-12-00536.1.

    • Search Google Scholar
    • Export Citation
  • Bond, N. A., M. F. Cronin, H. Freeland, and N. Mantua, 2015: Causes and impacts of the 2014 warm anomaly in the NE Pacific. Geophys. Res. Lett., 42, 34143420, https://doi.org/10.1002/2015GL063306.

    • Search Google Scholar
    • Export Citation
  • Bozkurt, D., O. L. Sen, Y. Ezber, B. Guan, M. Viale, and F. Caglar, 2021: Influence of African atmospheric rivers on precipitation and snowmelt in the Near East’s highlands. J. Geophys. Res. Atmos., 126, e2020JD033646, https://doi.org/10.1029/2020JD033646.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., C. Smith, and J. M. Wallace, 1992: An intercomparison of methods for finding coupled patterns in climate data. J. Climate, 5, 541560, https://doi.org/10.1175/1520-0442(1992)005<0541:AIOMFF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brönnimann, S., 2007: Impact of El Niño–Southern Oscillation on European climate. Rev. Geophys., 45, RG3003, https://doi.org/10.1029/2006RG000199.

    • Search Google Scholar
    • Export Citation
  • Cash, B. A., and N. J. Burls, 2019: Predictable and unpredictable aspects of U.S. West Coast rainfall and El Niño: Understanding the 2015/16 event. J. Climate, 32, 28432868, https://doi.org/10.1175/JCLI-D-18-0181.1.

    • Search Google Scholar
    • Export Citation
  • Chen, M., and A. Kumar, 2018: Winter 2015/16 atmospheric and precipitation anomalies over North America: El Niño response and the role of noise. Mon. Wea. Rev., 146, 909927, https://doi.org/10.1175/MWR-D-17-0116.1.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, https://doi.org/10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Deser, C., and M. Blackmon, 1995: On the relationship between tropical and North Pacific sea surface temperature variations. J. Climate, 8, 16771680, https://doi.org/10.1175/1520-0442(1995)008<1677:OTRBTA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Deser, C., I. R. Simpson, A. S. Phillips, and K. A. McKinnon, 2018: How well do we know ENSO’s climate impacts over North America, and how do we evaluate models accordingly? J. Climate, 31, 49915014, https://doi.org/10.1175/JCLI-D-17-0783.1.

    • Search Google Scholar
    • Export Citation
  • Dettinger, M. D., F. M. Ralph, T. Das, P. J. Neiman, and D. R. Cayan, 2011: Atmospheric rivers, floods and the water resources of California. Water, 3, 445478, https://doi.org/10.3390/w3020445.

    • Search Google Scholar
    • Export Citation
  • Dong, L., L. R. Leung, F. Song, and J. Lu, 2018: Roles of SST versus internal atmospheric variability in winter extreme precipitation variability along the U.S. West Coast. J. Climate, 31, 80398058, https://doi.org/10.1175/JCLI-D-18-0062.1.

    • Search Google Scholar
    • Export Citation
  • Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 19371958, https://doi.org/10.5194/gmd-9-1937-2016.

    • Search Google Scholar
    • Export Citation
  • Francis, D., K. S. Mattingly, M. Temimi, R. Massom, and P. Heil, 2020: On the crucial role of atmospheric rivers in the two major Weddell Polynya events in 1973 and 2017 in Antarctica. Sci. Adv., 6, eabc2695, https://doi.org/10.1126/sciadv.abc2695.

    • Search Google Scholar
    • Export Citation
  • Gao, Y., J. Lu, and L. R. Leung, 2016: Uncertainties in projecting future changes in atmospheric rivers and their impacts on heavy precipitation over Europe. J. Climate, 29, 67116726, https://doi.org/10.1175/JCLI-D-16-0088.1.

    • Search Google Scholar
    • Export Citation
  • Gershunov, A., T. Shulgina, F. M. Ralph, D. A. Lavers, and J. J. Rutz, 2017: Assessing the climate-scale variability of atmospheric rivers affecting western North America. Geophys. Res. Lett., 44, 79007908, https://doi.org/10.1002/2017GL074175.

    • Search Google Scholar
    • Export Citation
  • Gorodetskaya, I. V., M. Tsukernik, K. Claes, M. F. Ralph, W. D. Neff, and N. P. M. Van Lipzig, 2014: The role of atmospheric rivers in anomalous snow accumulation in East Antarctica. Geophys. Res. Lett., 41, 61996206, https://doi.org/10.1002/2014GL060881.

    • Search Google Scholar
    • Export Citation
  • Guan, B., and D. E. Waliser, 2015: Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. J. Geophys. Res. Atmos., 120, 12 51412 535, https://doi.org/10.1002/2015JD024257.

    • Search Google Scholar
    • Export Citation
  • Guan, B., D. E. Waliser, N. P. Molotch, E. J. Fetzer, and P. J. Neiman, 2012: Does the Madden–Julian oscillation influence wintertime atmospheric rivers and snowpack in the Sierra Nevada? Mon. Wea. Rev., 140, 325342, https://doi.org/10.1175/MWR-D-11-00087.1.

    • Search Google Scholar
    • Export Citation
  • Guan, B., D. E. Waliser, F. M. Ralph, E. J. Fetzer, and P. J. Neiman, 2016: Hydrometeorological characteristics of rain-on-snow events associated with atmospheric rivers. Geophys. Res. Lett., 43, 29642973, https://doi.org/10.1002/2016GL067978.

    • Search Google Scholar
    • Export Citation
  • Guirguis, K., A. Gershunov, T. Shulgina, R. E. S. Clemesha, and F. M. Ralph, 2019: Atmospheric rivers impacting Northern California and their modulation by a variable climate. Climate Dyn., 52, 65696583, https://doi.org/10.1007/s00382-018-4532-5.

    • Search Google Scholar
    • Export Citation
  • Henn, B., K. N. Musselman, L. Lestak, F. M. Ralph, and N. P. Molotch, 2020: Extreme runoff generation from atmospheric river driven snowmelt during the 2017 Oroville Dam spillways incident. Geophys. Res. Lett., 47, e2020GL088189, https://doi.org/10.1029/2020GL088189.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., R. F. Adler, M. M. Morrissey, D. T. Bolvin, S. Curtis, R. Joyce, B. McGavock, and J. Susskind, 2001: Global precipitation at one-degree daily resolution from multisatellite observations. J. Hydrometeor., 2, 3650, https://doi.org/10.1175/1525-7541(2001)002<0036:GPAODD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., Y. Kushnir, G. Ottersen, and M. Visbeck, 2003: An overview of the North Atlantic Oscillation. Climatic Significance and Environmental Impact, Geophys. Monogr., No. 134, Amer. Geophys. Union, 135.

  • Kim, H.-M., Y. Zhou, and M. A. Alexander, 2019: Changes in atmospheric rivers and moisture transport over the Northeast Pacific and western North America in response to ENSO diversity. Climate Dyn., 52, 7375–7388, https://doi.org/10.1007/s00382-017-3598-9.

    • Search Google Scholar
    • Export Citation
  • Kumar, A., and M. Chen, 2017: What is the variability in US west coast winter precipitation during strong El Niño events? Climate Dyn., 49, 27892802, https://doi.org/10.1007/s00382-016-3485-9.

    • Search Google Scholar
    • Export Citation
  • Lamjiri, M. A., M. D. Dettinger, F. M. Ralph, and B. Guan, 2017: Hourly storm characteristics along the U.S. West Coast: Role of atmospheric rivers in extreme precipitation. Geophys. Res. Lett., 44, 70207028, https://doi.org/10.1002/2017GL074193.

    • Search Google Scholar
    • Export Citation
  • Lavers, D. A., and G. Villarini, 2013: The nexus between atmospheric rivers and extreme precipitation across Europe. Geophys. Res. Lett., 40, 32593264, https://doi.org/10.1002/grl.50636.

    • Search Google Scholar
    • Export Citation
  • Lavers, D. A., and G. Villarini, 2015: The contribution of atmospheric rivers to precipitation in Europe and the United States. J. Hydrol., 522, 382390, https://doi.org/10.1016/j.jhydrol.2014.12.010.

    • Search Google Scholar
    • Export Citation
  • Lee, J., K. R. Sperber, P. J. Gleckler, C. J. W. Bonfils, and K. E. Taylor, 2019: Quantifying the agreement between observed and simulated extratropical modes of interannual variability. Climate Dyn., 52, 40574089, https://doi.org/10.1007/s00382-018-4355-4.

  • Lee, J., K. R. Sperber, P. J. Gleckler, K. E. Taylor, and C. J. W. Bonfils, 2021: Benchmarking performance changes in the simulation of extratropical modes of variability across CMIP generations. J. Climate, 34, 6945–6969, https://doi.org/10.1175/JCLI-D-20-0832.1.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., and Y. Qian, 2009: Atmospheric rivers induced heavy precipitation and flooding in the western U.S. simulated by the WRF regional climate model. Geophys. Res. Lett., 36, L03820, https://doi.org/10.1029/2008GL036445.

  • Li, C., W. Mei, and Y. Kamae, 2022: Variability and predictability of cold-season North Atlantic atmospheric river occurrence frequency in a set of high-resolution atmospheric simulations. Climate Dyn., 58, 24852500, https://doi.org/10.1007/s00382-021-06017-y.

    • Search Google Scholar
    • Export Citation
  • Li, Y., and N.-C. Lau, 2012: Impact of ENSO on the atmospheric variability over the North Atlantic in late winter—Role of transient eddies. J. Climate, 25, 320342, https://doi.org/10.1175/JCLI-D-11-00037.1.

    • Search Google Scholar
    • Export Citation
  • Ma, W., G. Chen, and B. Guan, 2020a: Poleward shift of atmospheric rivers in the Southern Hemisphere in recent decades. Geophys. Res. Lett., 47, e2020GL089934, https://doi.org/10.1029/2020GL089934.

    • Search Google Scholar
    • Export Citation
  • Ma, W., J. Norris, and G. Chen, 2020b: Projected changes to extreme precipitation along North American West Coast from the CESM large ensemble. Geophys. Res. Lett., 47, e2019GL086038, https://doi.org/10.1029/2019GL086038.

    • Search Google Scholar
    • Export Citation
  • Ma, W., G. Chen, Y. Peings, and N. Alviz, 2021: Atmospheric river response to Arctic sea ice loss in the Polar Amplification Model Intercomparison Project. Geophys. Res. Lett., 48, e2021GL094883, https://doi.org/10.1029/2021GL094883.

    • Search Google Scholar
    • Export Citation
  • Mattingly, K. S., T. L. Mote, and X. Fettweis, 2018: Atmospheric river impacts on Greenland Ice Sheet surface mass balance. J. Geophys. Res. Atmos., 123, 85388560, https://doi.org/10.1029/2018JD028714.

    • Search Google Scholar
    • Export Citation
  • McGowan, H., K. Borthwick, A. Schwartz, J. N. Callow, S. Bilish, and S. Browning, 2021: Atmospheric rivers: An overlooked threat to the marginal snowpack of the Australian Alps. J. Hydrometeor., 22, 25212532, https://doi.org/10.1175/JHM-D-20-0293.1.