Abstract

A major winter storm took place over Mexico during 7 to 11 March 2016, impacting 28 states and leaving four million families without power. Extensive agricultural damage and livestock deaths were also reported with widespread snow across central and northern Mexico. North of the border, this system resulted in record-breaking flooding and severe weather in Texas and Louisiana. The event was due to a trough that deepened and cut off over central Mexico with 500-hPa heights that were nine standard deviations below normal, well beyond previous records! Motivated by the societal impacts of this event, this study investigates factors that contributed to the extreme trough and influenced its predictability in forecast models. A strong El Niño provided the antecedent conditions, with enhanced tropical convection over the central Pacific, a strengthened subtropical anticyclone, and poleward Rossby wave dispersion. However, unlike past strong El Niños, the North Pacific preceding this event was characterized by significant synoptic-scale Rossby wave activity on the midlatitude jet stream including multiple wave packets tracking around the globe during February and March. The interaction of one of these packets with the subtropical anticyclone aloft resulted in a large anticyclonic wave break over the east Pacific, leading to the amplification of the downstream trough over Mexico. The ability of numerical weather prediction to capture this extreme trough is directly related to the predictability of the Rossby wave packet. These results are also discussed within the context of the relationship between El Niño, Rossby wave activity, and extreme events in western North America.

This study discusses the roles that a strong El Niño, the midlatitude wave guide, and long-lived Rossby wave packets played in a major winter storm that impacted Mexico in March 2016.

An extreme winter storm impacted Mexico in March 2016, in association with an exceptionally deep upper-level trough characterized by 500-hPa heights near nine standard deviations below normal (Fig. 1a). This storm had numerous detrimental societal and economic impacts. Much of Mexico was in a state of emergency resulting from severe wind and cold weather. An estimated four million people lost power, primarily in the states of Mexico and Puebla in the south-central region of the country. In addition, hundreds of injuries and fatalities were caused by the high wind and hypothermia. Schools were closed in 12 states. Major agricultural damage was declared in the states of Oaxaca, Sinaloa, Durango, San Luis Potosí, Tamaulipas, and Zacatecas. In the state of Sinaloa alone, nearly 50,000 acres of corn and wheat were harmed by the wind and heavy precipitation. The event was also responsible for harmful environmental impacts. Freezing cold led to a mortality rate of over 30% of butterflies in colonies of the Monarch Butterfly Biosphere Reserve (Brower et al. 2017). Strong winds with this storm toppled thousands of trees, destroying shielded sanctuaries and scouring out the microclimate that butterflies had relied on in the past cold air outbreaks.

Fig. 1.

(a) Contours of 500-hPa heights at 1200 UTC 10 Mar 2016 from ERA-Interim. The standard deviation of the anomaly from the climatological mean is given by the color shading. (b) Daily-mean 500-hPa heights over central Mexico indicated by the white box on the map (17.5°–22.5°N, 100°–105°W) from NCEP–NCAR reanalysis for 1948–2015. Light gray fill marks the 10th–90th percentile range and dark gray fill marks the 25th–75th percentile range. Red, green, and blue lines mark the daily maximum, median, and minimum, respectively. The black line marks the 2015/16 daily heights.

Fig. 1.

(a) Contours of 500-hPa heights at 1200 UTC 10 Mar 2016 from ERA-Interim. The standard deviation of the anomaly from the climatological mean is given by the color shading. (b) Daily-mean 500-hPa heights over central Mexico indicated by the white box on the map (17.5°–22.5°N, 100°–105°W) from NCEP–NCAR reanalysis for 1948–2015. Light gray fill marks the 10th–90th percentile range and dark gray fill marks the 25th–75th percentile range. Red, green, and blue lines mark the daily maximum, median, and minimum, respectively. The black line marks the 2015/16 daily heights.

The impending hostile environment was led by a cold front stretching across northwest Mexico on 7 March. The synoptic evolution and societal impacts during the days that followed are depicted in Fig. 2. Temperatures plummeted from near 20°C to below freezing with winds gusting to 50 mph. For the next three days, most of central Mexico averaged between 6° and 10°C below normal (Fig. 3a) following what was a temperate winter (Fig. 3c). Sleet and snow was reported over the Sierra Madre Occidental and as far south as Guadalajara, and frost was reported in Mexico City. Along the west coast of Mexico, snow was recorded for the first time in northern Sinaloa and western Durango with record wind and low temperatures reaching southward to as far as Chiapas and Oaxaca. Downstream of the trough, deep-layer moisture transport from the Gulf of Mexico fueled heavy rain in strong to severe thunderstorms from Texas to Louisiana, Arkansas, and Mississippi. Because of the cutoff nature of the trough, this regime was stagnant between 8 and 12 March, during which time widespread rainfall over one foot (Fig. 3b) resulted in devastating flooding (Breaker et al. 2016).

Fig. 2.

Positions of the surface cold front at 2100 UTC 8 Mar, 0900 UTC 9 Mar, and 1800 UTC 10 Mar 2016. Inset figures and annotations point to locations of societal impacts from the cold air outbreak.

Fig. 2.

Positions of the surface cold front at 2100 UTC 8 Mar, 0900 UTC 9 Mar, and 1800 UTC 10 Mar 2016. Inset figures and annotations point to locations of societal impacts from the cold air outbreak.

Fig. 3.

Summary of temperatures and precipitation during the March storm using North American Regional Reanalysis (Mesinger et al. 2006): (a) 3-day 2-m temperature anomaly for 9–11 Mar 2016, (b) 5-day accumulated precipitation for 8–12 Mar. Summary of the winter of 2015/16 using ERA-Interim (Dee et al. 2011): (c) temperature anomaly for December 2015 through February 2016, (d) precipitation percent of normal for December 2015 through February 2016. All anomalies use a 1981–2010 climatology.

Fig. 3.

Summary of temperatures and precipitation during the March storm using North American Regional Reanalysis (Mesinger et al. 2006): (a) 3-day 2-m temperature anomaly for 9–11 Mar 2016, (b) 5-day accumulated precipitation for 8–12 Mar. Summary of the winter of 2015/16 using ERA-Interim (Dee et al. 2011): (c) temperature anomaly for December 2015 through February 2016, (d) precipitation percent of normal for December 2015 through February 2016. All anomalies use a 1981–2010 climatology.

The origin of the bitter cold beneath the deep upper-level trough was investigated through backward trajectories that show this air mass primarily originated over the subtropical and midlatitude Pacific (Fig. 4). These trajectories suggest that this event had a different synoptic character than many of the cold air outbreaks that impact Mexico, which often originate in Canada with a strong anticyclone and move rapidly along the lee side of the Rockies because of interaction of the terrain with the evolving synoptic flow (Colle and Mass 1995; Schultz et al. 1998). These leeside cold air outbreaks are often referred to as nortes (Magaña et al. 2003) and are discussed, for example, in Henry (1979), Reding (1992), Schultz et al. (1997), Pérez et al. (2014), and Luna-Niño and Cavazos (2018). The frequency and structure of cold surge activity varies with El Niño–Southern Oscillation (ENSO) (Magaña et al. 2003; Magaña and Ambrizzi 2005) with less cold surges during El Niño. This winter storm took place during the strong 2015/16 El Niño (e.g., L’Heureux et al. 2017; Dole et al. 2017) that will be described in the next section.

Fig. 4.

Backward trajectories for 120 h starting from grid points between 1 km above ground level and 500 hPa, over central Mexico at 1200 UTC 10 Mar 2016 using NOAA ARL’s HYSPLIT program (Stein et al. 2015). The input data were from the North American Regional Analysis. Each point is 1 h, and the color refers to the potential temperature.

Fig. 4.

Backward trajectories for 120 h starting from grid points between 1 km above ground level and 500 hPa, over central Mexico at 1200 UTC 10 Mar 2016 using NOAA ARL’s HYSPLIT program (Stein et al. 2015). The input data were from the North American Regional Analysis. Each point is 1 h, and the color refers to the potential temperature.

A FLAVOR OF STRONG EL NIÑO.

The winter of 2015/16 featured one of the strongest El Niños on record (e.g., Blunden and Arndt 2016; L’Heureux et al. 2017; Santoso et al. 2017; Dole et al. 2017). Seasonal predictions of sensible weather made in the fall of 2015 were based heavily on this exceptional ENSO state. As discussed in Dole et al. (2017), the tendency for heavy rainfall to occur in California in association with strong El Niños in the past brought specific interest to this event given the region’s persistent historic drought. However, the drought would not be quelled this winter (Fig. 3d). One wild card noted in seasonal outlooks was that SSTs across the North Pacific did not resemble the typical configuration of a positive Pacific decadal oscillation (PDO) that would normally accompany such an authoritative tropical regime (Newman et al. 2003). Above-average SSTs controlled much of the subtropical and midlatitude ocean, especially in a region eloquently labeled the “warm blob” between Hawaii and Baja California (e.g., Peterson et al. 2015). As a result, the meridional SST gradient was less than typically present with a strong El Niño.

Strong El Niños feature enhanced tropical convection in the central and eastern Pacific, associated with a dominant Hadley cell and attendant twin anticyclones at the subtropical tropopause (Ropelewski and Halpert 1987; Oort and Yienger 1996). Through conservation of angular momentum, the upper-level poleward branch of the Hadley cell powers an enhanced subtropical jet from the central Pacific across tropical/subtropical North America (Rosen et al. 1984). The latent heat release associated with tropical convection reduces stability aloft and pushes the tropopause upward along with a decrease in potential vorticity (PV). Strong divergence above the convection advects the low-PV air out of the tropics. As these parcels trace poleward they encounter a northward-directed PV gradient as a result of increasing planetary vorticity by latitude, and can become large negative PV perturbations superimposed on the background flow field (Sardeshmukh and Hoskins 1988). In the presence of a jet stream with a significantly tightened PV gradient within the background flow, high-wavenumber (short wavelength) Rossby waves, and their associated synoptic weather systems, are channeled downstream along the spine of the PV gradient maximum (Hoskins and Ambrizzi 1993). Low-wavenumber Rossby wave activity emanating from the initial perturbation is free to propagate farther poleward (e.g., Fig. 2 in Hoskins and Ambrizzi 1993).

In past strong El Niño events, the subtropical jet was fast and narrow yielding a large background PV gradient (Held and Hou 1980; Bell and Halpert 1998). This PV distribution means that the jet acts as a filter in which only the lowest-wavenumber Rossby wave disturbances triggered from tropical convection can reach beyond the subtropics, essentially shadowing higher latitudes from the shorter wavelengths. The most common manifestation of this Rossby wave dispersion is the Pacific–North American (PNA) pattern (Wallace and Gutzler 1981), in which a planetary-scale wave response emanates northeastward from enhanced convection over the equatorial Pacific. As a result of this filtering effect, the winters with an augmented subtropical jet would be dominated by low-wavenumber perturbations in the midlatitudes.

In stark contrast to this painting of a canonical strong El Niño, the winter of 2015/16 was characterized by a weaker thermal gradient in SSTs due to the warm waters across portions of the subtropics and midlatitudes. Utilizing two ensembles of climate simulations, Quan et al. (2018) note the importance of differences in the boundary conditions outside of the ENSO core region and a weaker and poleward-shifted upper-level low anomaly over the North Pacific compared to the typical ENSO teleconnection. We suggest that the wider expanse of warm waters may link to their conclusions, in addition to producing a wider and weakened Hadley cell yielding a less-dominant subtropical jet. These shifts can all contribute to a weaker poleward-directed PV gradient in the subtropics, and as a result, Rossby wave activity and related storm tracks and patterns of rainfall could be expected to have different behavior in the subtropics and midlatitudes than past strong El Niño winters. Flow patterns through 2015/16 were thus likely to be characterized by a mix of both low- and high-wavenumber activity, and greater impact on significant weather coming from internal variability. The importance of this internal variability in determining the relationship between ENSO and rainfall on the west coast of North America has been expressed by Deser et al. (2018).

To examine differences in the behavior of Rossby wave activity dependent on the structure of the Pacific jets and PV field, we treat the atmosphere as a refractive medium that guides the path of Rossby wave activity of varying wavelengths. Similar to geometric optics, the group velocity of Rossby wave activity can be described as following rays that track and bend according to an index of refraction, n (Hoskins and Karoly 1981; Hoskins and Ambrizzi 1993). Where n is a real number, the basic-state flow supports the existence of Rossby wave activity of a given zonal wavenumber, k. An imaginary n, or negative n2, identifies where this wave activity is evanescent, and at n2 equal to zero this wave activity is reflected or absorbed (Held 1983). A waveguide for waves of a given k is thus defined by a band of positive n2 flanked to the north and south by negative n2 (e.g., Fig. 2 in Hoskins and Ambrizzi 1993). Originally from Matsuno (1970) and later derived by O’Rourke and Vallis (2016), n2 can be written as

 
n2=1u¯cq¯ϕk2cos2(ϕ),
(1)

where u is the zonal wind, c is the zonal component of the Rossby wave phase velocity, ∂q/∂ϕ is the meridional gradient in PV, and the overbar indicates the basic state, which in this study is defined by a 1-month time average. The zonal wavenumber of Rossby wave activity, k, is a constant for which n2 is evaluated. The wave train during February and March 2016 featured circumglobal wavenumbers between 5 and 7 (which can be visually confirmed with the Hovmöller diagram in Fig. 5a). Specifically, wavenumber 5 is at an important intersection where the wave is long enough to have a slow phase speed, but short enough to have dynamical significance for synoptic-scale weather. Additionally, it has been demonstrated to be associated with extreme ridges near the west coast of North America (Teng and Branstator 2017). For these reasons, in this study we evaluate n2 for stationary zonal wavenumber-5 wave activity in 2016 and compare to past years using National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) reanalysis.

Fig. 5.

(a) Hovmöller diagram of 6-hourly 250-hPa meridional wind from NCEP–NCAR reanalysis, averaged between 20°–50°N, for February and March 2016. The dashed green line marks the primary RWP during this period. The green circles denote two significant weather events directly connected to the RWP: 1) a tornado outbreak in the southeast United States on 23–24 Feb, and 2) the historic Mexico trough on 7–11 Mar. (b) The 250-hPa perturbation streamfunction (color fill) and wave activity flux (vectors) for 6–10 Mar 2016.

Fig. 5.

(a) Hovmöller diagram of 6-hourly 250-hPa meridional wind from NCEP–NCAR reanalysis, averaged between 20°–50°N, for February and March 2016. The dashed green line marks the primary RWP during this period. The green circles denote two significant weather events directly connected to the RWP: 1) a tornado outbreak in the southeast United States on 23–24 Feb, and 2) the historic Mexico trough on 7–11 Mar. (b) The 250-hPa perturbation streamfunction (color fill) and wave activity flux (vectors) for 6–10 Mar 2016.

The frequencies of n2 above and below zero in a given subset climatology provides some insight into the probability that Rossby wave activity of wavenumber 5 can be supported and guided eastward. The influence of ENSO on this metric is examined using a composite difference (El Niño minus La Niña) in the frequency of n2 < 0 (Fig. 6a). From this analysis, there is a large region of significant positive difference in the central equatorial Pacific that is a product of a shifted Walker circulation, where winds tied to large-scale convective outflow in the upper levels of the atmosphere are westerly during La Niña and easterly during El Niño. Over the Atlantic the subtropical jet waveguide is strongly favored during El Niño (Fig. 6a, red shading), and over the contiguous United States the polar jet waveguide is favored during La Niña (Fig. 6a, blue shading). Meanwhile, the differences are generally not statistically significant over the extratropical Pacific, suggesting little influence from ENSO on the pathways of Rossby wave activity upstream of North America. This lack of significant association within ENSO states lends itself to internal variability (Deser et al. 2017) leading to different time-mean patterns between ENSO cases.

Fig. 6.

Spatial distribution of the frequency of negative refractive index for zonal wavenumber 5. (a) Difference between top 10 El Niño vs La Niña years in January–March. Stippling indicates significance at 95% confidence. Negative values (shaded red) correspond to wavenumber-5 activity favored during El Niño, whereas positive values (shaded blue) indicate wave activity favored during La Niña. (b) Composite of strong El Niños (1958, 1973, 1983, 1992, 1998) for January–March. Low values, near zero, correspond with conditions that nearly always sustain wavenumber-5 activity. High values, near one, correspond with conditions that nearly never sustain wavenumber-5 activity. Regions with values near 0.5 indicate variability within the climatology.

Fig. 6.

Spatial distribution of the frequency of negative refractive index for zonal wavenumber 5. (a) Difference between top 10 El Niño vs La Niña years in January–March. Stippling indicates significance at 95% confidence. Negative values (shaded red) correspond to wavenumber-5 activity favored during El Niño, whereas positive values (shaded blue) indicate wave activity favored during La Niña. (b) Composite of strong El Niños (1958, 1973, 1983, 1992, 1998) for January–March. Low values, near zero, correspond with conditions that nearly always sustain wavenumber-5 activity. High values, near one, correspond with conditions that nearly never sustain wavenumber-5 activity. Regions with values near 0.5 indicate variability within the climatology.

Looking at a climatology of strong El Niño winters, we see an indecisive Pacific as well (Fig. 6b). Large parts of the subtropics and midlatitudes with frequencies between 0.3 and 0.7 indicate areas that have seen both positive and negative values of n2, where there is diversity in the location of waveguides. This variability means that even within the subset of strong El Niños, the path of Rossby wave activity out of the tropics and in the midlatitudes can vary yielding a range of influence in North America (see the sidebar for an examination of low-frequency Rossby wave behavior in past strong El Niños). The refractive index analysis demonstrates how the midlatitude Rossby wave response associated with ENSO can vary beyond the canonical pattern. Indeed there can be considerable variability in the wintertime extratropical patterns and resulting sensible weather that coincides with the same low-frequency tropical forcing. While it may be tempting to dismiss this as noise on top of a canonical ENSO regime, this dismissal rejects the critical interactions that occur between the prevalent ENSO forcing, the distribution of basic-state PV, and midlatitude Rossby wave dynamics. Now we ask what are the connections between these players and how they shaped an unusual El Niño winter including the historic winter storm in Mexico and the lack of extreme rainfall in California.

ROSSBY WAVE BEHAVIOR IN PAST EL NIÑOS

This study explores the importance of the nature of the Rossby waveguides over the Pacific in determining how the shift in the location of deep tropical convection associated with El Niño impacts weather over North America. Since our investigation is based primarily on analysis of the 2016 El Niño event, we explore this concept for six strong El Niño years using streamfunction and velocity potential anomalies, and WAF (Takaya and Nakamura 1997) for stationary eddies during the month of February ( Fig. SB1). When applied to a time-averaged field such as this, WAF tells us about the paths of slowly evolving Rossby wave activity. There are certain pattern commonalities in proximity to the anomalous tropical convective forcing, such as the anomalous subtropical anticyclones seen in the streamfunction on both sides of the equator, a characteristic of an enhanced Hadley circulation (Schneider 1987). However, subseasonal patterns of troughs and ridges across the midlatitudes still exhibit marked variability. The strength and position of jet streams influence the paths of Rossby waves (Wirth et al. 2018) by acting as waveguides, and wave activity in the midlatitudes leads to variability within this El Niño climatology. For example, Caballero and Anderson (2009) demonstrated that Hadley cell strength coincides with midlatitude wave activity that gets reabsorbed into the tropical central Pacific. This meridional flux can be seen at varying magnitudes in past cases (Fig. SB1). The general pattern of WAF also varies as, for example, February 2016 featured strong zonally oriented wave activity flux with a higher-wavenumber pattern across the North Pacific. Rossby wave activity emanated from a source region in East Asia, continually amplifying a semipermanent ridge over the west Pacific. One consequence of this pattern was a shift of the jet away from California, with ridging and anomalously dry conditions, consistent with the findings in Quan et al. (2018). This is in contrast to observations from Jong et al. (2016) that El Niño’s influence on California rainfall was strongest to the south and in late winter. Every other strong El Niño year in Fig. SB1 had an extension of the jet into the west coast of North America evident by the meridional gradient in the streamfunction anomaly, along with negative velocity potential anomalies overhead supportive of lift and precipitation. Thus, the interaction between propagating midlatitude Rossby wave activity, changes in the location of ENSO-related tropical convection, and the nature of the Pacific waveguides associated with the subtropical and polar jets interact to produce considerable variability between El Niño events.

Fig. SB1.

The 250-hPa analysis for the February of strong El Niño years. The contours show streamfunction anomalies at intervals of 5 × 106 m2 s–1, with bold lines every 20 × 106 m2 s–1. The velocity potential anomaly is given by the color shading. Vectors indicate wave activity flux.

Fig. SB1.

The 250-hPa analysis for the February of strong El Niño years. The contours show streamfunction anomalies at intervals of 5 × 106 m2 s–1, with bold lines every 20 × 106 m2 s–1. The velocity potential anomaly is given by the color shading. Vectors indicate wave activity flux.

During February 2016, the midlatitude waveguide was well defined in the western Pacific and shifted poleward from climatology (Fig. 7a). East of the date line, the waveguide bent equatorward, and merged with the subtropical waveguide in the eastern Pacific into Mexico. We illustrate the theoretical path of Rossby wave activity using the ray tracing framework established in Hoskins and Karoly (1981). The group velocity for barotropic Rossby waves is derived as cg = ∂ω/K from the dispersion relation, ω=u¯kq¯(k/|K|2), where K=ki^+lj^ is the wavenumber vector. Ray paths are constructed by integrating forward dx/dt = cgx and dy/dt = cgy for a given constant k and l that is evaluated at each time step. Rossby wave activity generated or reinvigorated over East Asia and the west Pacific baroclinic zone travels eastward in the midlatitude waveguide, and splits near 210°E, with shorter waves turning equatorward into the subtropics (Fig. 7b). Wave sources in the tropics result in planetary-scale wave activity tracking poleward across the eastern Pacific into Canada (Fig. 7c, low wavenumber, yellow and orange lines). Synoptic-scale wave activity meanwhile gets channeled within the subtropical waveguide (Fig. 7c, high wavenumber, red and purple lines), merging with wave activity from the midlatitudes into Mexico. It is the nature of the midlatitude waveguide and subtropical waveguide, and the merger of the two in the eastern Pacific as described in Wirth et al. (2018), that is key in producing a favorable environment for long-lasting groups of high-amplitude Rossby waves known as Rossby wave packets (RWPs). The role of RWPs in the extreme weather events in 2016 will be discussed in the next section.

Fig. 7.

Basic-state analysis for preferred Rossby wave activity propagation in February 2016, using the average 250-hPa winds for the month. (a) Refractive index for zonal wavenumber 5. (b) Rossby ray traces integrated forward every 60 s for 5 days following the equations of Hoskins and Karoly (1981), for zonal wavenumbers 2–7 as indicated by their color. Rays are initialized from a matrix of points every 2.5° latitude and 5° longitude in the midlatitude western Pacific. (c) As in (b), but for rays initialized in the tropical Pacific.

Fig. 7.

Basic-state analysis for preferred Rossby wave activity propagation in February 2016, using the average 250-hPa winds for the month. (a) Refractive index for zonal wavenumber 5. (b) Rossby ray traces integrated forward every 60 s for 5 days following the equations of Hoskins and Karoly (1981), for zonal wavenumbers 2–7 as indicated by their color. Rays are initialized from a matrix of points every 2.5° latitude and 5° longitude in the midlatitude western Pacific. (c) As in (b), but for rays initialized in the tropical Pacific.

ROSSBY WAVE ACTIVITY.

During the winter of 2015/16, there were two extended periods of elevated Rossby wave activity with multiple RWPs circling the Northern Hemisphere midlatitudes. The first period, while not a focus of this study, is briefly described to introduce the role of RWPs in high-impact events this winter. This period featured an RWP tracking from the Pacific across the Western Hemisphere, during mid-December through early January. The RWP was responsible for the amplification of ridging over the northeast United States around Christmas, causing record-breaking warmth. A downstream trough amplified over the North Atlantic, subsequently breaking cyclonically poleward and the wave train culminated in a highly anomalous anticyclone over the Kara Sea by the first week of January. The amplification of this high-latitude ridging resulted in a breakdown of the tropospheric polar vortex, manifested in the Arctic Oscillation (AO) tumbling from a daily record high over +4 sigma in the third week of December to –4 sigma by mid-January. Record warmth flooded the Arctic in association with this ridge (Cullather et al. 2016). Meanwhile the negative annular mode preceded a crippling blizzard in the mid-Atlantic United States on 22–24 January. Widespread one to three feet of snow and blizzard conditions paralyzed travel, knocked out power to more than half a million people, and resulted in 55 fatalities (Uccellini 2016).

The second period of elevated Rossby wave activity featured two RWPs circling the midlatitude Northern Hemisphere concurrently from the end of January through the first half of March. The two RWPs were the catalysts for multiple high-impact weather events through February and March (Fig. 5a), including a tornado outbreak in the southeast United States on 23–24 February. A total of 65 tornadoes were reported over the two days, resulting in seven fatalities [Storm Prediction Center (SPC) tornado database; Schaefer and Edwards 1999]. The same RWP maintained coherence while circling the globe once more through the first week of March, reaching the eastern Pacific and North America by 7 March.

At this point in March, a Madden–Julian oscillation (MJO) wave propagating into the Western Hemisphere combined with low-frequency ENSO forcing to enhance the anomalous tropical convection over the eastern Pacific (Blunden and Arndt 2016). The seasonal and subseasonal forcing drove broad twin anticyclones on either side of the equator. The extratropical RWP reenergized (Fig. 5b) as it reached the Pacific waveguide (Fig. 7). The RWP continued into the Western Hemisphere and turned southeastward as illustrated by the positive (southwest to northeast) tilt of the trough axis in Fig. 5b, consistent with the pathway defined by the ray traces in Fig. 7b. As the RWP moved downstream, it constructively interfered with the large-scale anticyclone in the subtropical North Pacific, resulting in both amplification and shortening of the wavelength of the ridge. Subsequent anticyclonic wave breaking toward Mexico accompanied the passage of the RWP downstream. On the northeast edge of the ridge, air dragged from the cyclonic shear side of the jet carved a trough along the U.S. West Coast toward Baja California (Fig. 5b).

With an upper-level ridge anchored over the eastern Pacific in association with tropical forcing, and the RWP turning equatorward toward Mexico between 8 and 10 March (Fig. 5b), the downstream trough only continued amplifying southward. By 10 March, this trough had cut off over central Mexico with geopotential heights in the center of the 500-hPa low below 5,600 m. The anomaly extremum, centered near 20°N, 103°W, dipped to around nine standard deviations below normal (Fig. 1a). The depth of this low far exceeded the previous all-time record 500-hPa height minimum over central Mexico in NCEP–NCAR reanalysis (Fig. 1b). The RWP continued eastward, amplifying a ridge downstream over the eastern United States.

These RWPs can be diagnosed using the wave activity flux (WAF) from Takaya and Nakamura (2001) as demonstrated in Wolf and Wirth (2017). We take this approach to qualitatively examine propagation of the RWP in the beginning of March (Fig. 5b). To quantitatively assess the aggregate measure of RWP frequency and magnitude during this period of time and past years, the WAF for stationary eddies (Takaya and Nakamura 1997) is evaluated where the background winds are westerly and averaged over a 30-day moving window, and then averaged spatially across the midlatitude Pacific and the Northern Hemisphere. These measures were among the highest on record observed during this period according to the climatology calculated from the NCEP–NCAR reanalysis from 1950 to 2016 (Fig. 8). Past strong El Niños (Figs. 8a,b, red lines) exhibited above-normal magnitudes of WAF in the Northern Hemisphere and Pacific in October, which 2016 mimicked. Moving into December, these years regress toward the middle of the climatological envelope (Figs. 8a,b, gray lines). Decreasing wave activity in January and early February is noted across the entire climatology, which is consistent with midwinter suppression in the amplitude of baroclinic waves (Nakamura 1992). The strong Niño years tended to exhibit an even more pronounced minimum in the North Pacific (Fig. 8b), which was then followed by consistently below-normal WAF in the latter half of winter. In stark contrast, the midwinter suppression in 2016 was cut short and replaced with a period of wave activity at record high levels in the Northern Hemisphere (Fig. 8a, blue line) and near records over the Pacific as well (Fig. 8b, blue line). Average WAF over the Pacific was around 50% higher than any of the strong Niño cases. None of these past cases featured any exceptional levels of WAF in the latter half of winter comparable to February 2016.

Fig. 8.

The 30-day running-mean zonal component of wave activity flux at 250 hPa averaged over (a) the Northern Hemisphere, and (b) the midlatitude North Pacific (20°–50°N, 150°E–120°W) calculated from daily NCEP–NCAR reanalysis. Each year is plotted in gray (from 1 Jul to 30 Jun) from 1950 through 2016. Strong El Niños in January–March (1958, 1973, 1983, 1992, 1998) are plotted in red; 2015/16 is plotted in blue.

Fig. 8.

The 30-day running-mean zonal component of wave activity flux at 250 hPa averaged over (a) the Northern Hemisphere, and (b) the midlatitude North Pacific (20°–50°N, 150°E–120°W) calculated from daily NCEP–NCAR reanalysis. Each year is plotted in gray (from 1 Jul to 30 Jun) from 1950 through 2016. Strong El Niños in January–March (1958, 1973, 1983, 1992, 1998) are plotted in red; 2015/16 is plotted in blue.

PREDICTABILITY.

While the overall performance of numerical weather prediction is improving (e.g., Thorpe 2004; Shapiro et al. 2010), the ability to accurately predict individual extreme events is of highest consequence. This skill in predicting extremes is a concern not only because of the anomalous nature of the atmosphere at that time, but also the increased importance of the forecast to the public. Rossby wave packets have been shown to be a key to both high-impact weather events (Shapiro and Thorpe 2004; Wirth and Eichhorn 2014), as well as influence on medium- to long-range predictability (Grazzini and Vitart 2015). For the 3-month period from January through March, both the Global Forecast System (GFS) and European Centre for Medium-Range Weather Forecasts (ECMWF) had their best performances during the month of February (Fig. 9), coinciding with the time of the elevated RWP activity. In March, elevated forecast skill (Fig. 9) occurs during the passage of the strong RWP responsible for the Mexico trough (Fig. 5a), whereas weaker RWPs crossing the Pacific around 1 March and again around 15 March (Fig. 5a) are associated with reduced forecast skill, especially by the middle of the month.

Fig. 9.

(top) GFS and (bottom) ECMWF 500-hPa height anomaly correlation coefficient (ACC) over the North Pacific (10°–60°N, 120°E–270°W), standardized by lead time against the average and standard deviation of ACC during the 3-month period. Positive (negative) values indicate a forecast performance that is better (worse) than the 3-month average for the given lead time.

Fig. 9.

(top) GFS and (bottom) ECMWF 500-hPa height anomaly correlation coefficient (ACC) over the North Pacific (10°–60°N, 120°E–270°W), standardized by lead time against the average and standard deviation of ACC during the 3-month period. Positive (negative) values indicate a forecast performance that is better (worse) than the 3-month average for the given lead time.

The coherence of the wave train across the Pacific is a key factor in the forecast around the midlatitude Northern Hemisphere during this period. The ensemble mean and spread for the meridional wind at 250 hPa are shown in Fig. 10 for the National Oceanic and Atmospheric Administration’s (NOAA) Global Ensemble Forecast System (GEFS) and the ECMWF Ensemble Prediction System (EPS) for lead times up to 10 days. These forecast systems show striking skill in predicting aspects of the RWP even at the 10-day lead time. For example, the trough over Mexico is consistently evident in both the GEFS and EPS along with a downstream ridge over the eastern United States and a trough over the western Atlantic. Both ensemble means depict a strong trough over Mexico between day 8 and day 10 (bottom of Fig. 10), and already imply a potentially extreme event (Fig. 11). In addition, seasonal ensemble systems depicted the trough beyond 10 days as well (not shown). This skill in capturing the general pattern as shown in Fig. 10 is in agreement with a study from Grazzini and Vitart (2015) indicating increased predictability is associated with long-lived RWPs. However, there is still uncertainty contained within the medium-range forecast. The ensemble spread of the meridional wind shown by the contours in Fig. 10 is maximized between the northerly and southerly phases of the wave over North America (e.g., see EPS between 96 and 240 h). This placement indicates differences in the phase location of the troughs and ridges within the RWP. All waves are amplified within an RWP regardless of phase. If two sets of waves, such as a forecast and an analysis, within the same RWP have phase differences, errors from those differences will also amplify within the packet.

Fig. 10.

Ensemble mean (fill) and spread (contoured every 5 m s–1) of 250-hPa meridional wind forecasts from the (left) GEFS and (right) EPS for 1200 UTC 10 Mar 2016. Lead time starts at 0 h at the top, and incrementally increases by 48 h, to 240 h at the bottom.

Fig. 10.

Ensemble mean (fill) and spread (contoured every 5 m s–1) of 250-hPa meridional wind forecasts from the (left) GEFS and (right) EPS for 1200 UTC 10 Mar 2016. Lead time starts at 0 h at the top, and incrementally increases by 48 h, to 240 h at the bottom.

Fig. 11.

Ensemble forecasts from GEFS (red) and EPS (blue) for 500-hPa heights over central Mexico valid 1200 UTC 10 Mar 2016. The boxes indicate the 25th–75th percentile range, and the whiskers indicate the full range. The horizontal dashed line indicates the climatology from ERA-Interim and the solid line indicates the 1200 UTC 10 Mar 2016 verification from ERA-Interim.

Fig. 11.

Ensemble forecasts from GEFS (red) and EPS (blue) for 500-hPa heights over central Mexico valid 1200 UTC 10 Mar 2016. The boxes indicate the 25th–75th percentile range, and the whiskers indicate the full range. The horizontal dashed line indicates the climatology from ERA-Interim and the solid line indicates the 1200 UTC 10 Mar 2016 verification from ERA-Interim.

After recognition of the existence of the trough, the next question is its amplitude. We hypothesize that the depth of the trough over Mexico is tied to the intensity of the RWP moving across the Pacific. Using both the GEFS and EPS ensembles, we investigate the significance of this link by the sensitivity of the 500-hPa trough to 250-hPa winds. The depth of the trough is given by the negative of 500-hPa heights (such that a higher number indicates a deeper trough) in the central Mexico domain outlined in Fig. 1a. The sensitivity of the trough to the RWP is represented by the correlation of 250-hPa meridional wind at each grid point to the negative of 500-hPa heights over central Mexico, across the members of the ensemble (Fig. 12). The GEFS in particular demonstrates strong dependence of the Mexico trough on the upstream RWP as indicated by the train of significant correlation of alternating signs across the Pacific. In both the EPS and GEFS, positive correlations over Mexico are shifted west from the ensemble mean southerlies downstream of the trough. These phase-shifted correlations suggest a sensitivity of the depth of the trough to both the longitude of the trough axis, as well as its half wavelength. Following the positions of correlation extrema in Fig. 12, ensemble members with the lowest geopotential heights over Mexico were a product of a shorter-wavelength trough, as indicated by less distance between the correlation extrema than between the ensemble mean wind extrema. The same comparison also indicates the members with a deeper trough anchored the trough farther west and closer to the eastern Pacific subtropical ridge. The sensitivity to phase is also important given that the greatest ensemble spread was associated with the phase of synoptic features across the Pacific (Fig. 10).

Fig. 12.

Ensemble sensitivity for the depth of the 500-hPa trough (negative of geopotential height) over central Mexico to 250-hPa meridional winds in the day-8 forecast from (top) GEFS and (bottom) EPS. The correlation at each grid point is given by the color fill, with only the values above the 95% confidence threshold plotted. Ensemble mean forecast 250-hPa meridional winds are contoured every 5 m s–1, excluding the zero wind line.

Fig. 12.

Ensemble sensitivity for the depth of the 500-hPa trough (negative of geopotential height) over central Mexico to 250-hPa meridional winds in the day-8 forecast from (top) GEFS and (bottom) EPS. The correlation at each grid point is given by the color fill, with only the values above the 95% confidence threshold plotted. Ensemble mean forecast 250-hPa meridional winds are contoured every 5 m s–1, excluding the zero wind line.

To directly and quantitatively evaluate the handling of the RWP in the models, we use the zonal component of WAF area averaged across the midlatitude Pacific domain. Figure 13 depicts the interquartile range (IQR), defined as the range between the 25th and 75th percentiles, of the GEFS forecast area-averaged WAF as a function of lead time. The passage of RWPs is represented by the sinusoidal rise and fall in the averaged WAF. At 10-day lead time, the IQR is successful at capturing the verified WAF during the passage of the RWP between 7 and 10 March (Fig. 13). However, in the medium-range forecast, spread decreases substantially as indicated by the length of the bars shrinking at subsequent lead times. While the medium-range GEFS appropriately recognizes the passage of the RWP in early March, the amplitude is significantly underdone. At only 3-day lead time, the amplitude of the RWP in the GEFS mean is about 25% lower than the analysis.

Fig. 13.

GFS ensemble IQR of zonal WAF averaged over the midlatitude North Pacific (20°–50°N, 150°E–120°W). End points of each vertical line indicate the 25th and 75th percentiles of the ensemble forecast. Color indicates the lead time of the forecast. Black horizontal lines indicate the analysis.

Fig. 13.

GFS ensemble IQR of zonal WAF averaged over the midlatitude North Pacific (20°–50°N, 150°E–120°W). End points of each vertical line indicate the 25th and 75th percentiles of the ensemble forecast. Color indicates the lead time of the forecast. Black horizontal lines indicate the analysis.

As a result of the relatively weaker modeled RWP (Fig. 13), the GEFS forecasts of 500-hPa heights over central Mexico indicate a trough that is less amplified than the verification and exhibit less spread at day-6 to day-2 lead time (Fig. 11). We see that the ensemble systems captured the existence of a trough over Mexico at day-8 to day-10 lead time (Fig. 10), with the IQR of geopotential heights extending to around four standard deviations below normal in both forecast systems (Fig. 11). The GEFS forecast lower heights than the EPS from day 10 to day 5. Despite the consensus for a trough, the depth and equatorward extent of the trough do not begin to be captured until day 5 for the GEFS, and day 4.5 for the EPS. The EPS tends to lag the GEFS by 12 h in trends, though by day 4 consistently forecast lower heights than the GEFS for the remainder of forecast runs (Fig. 11). In summary, the expectation of an extreme event for central Mexico had approximately five days of lead time.

SUMMARY.

The winter of 2015/16 will be known for many things, including one of the strongest El Niños in recorded history. Additionally, it will be known for unexpected El Niño behavior. For example, the much-needed plentiful rainfalls projected by seasonal forecasts did not occur over California and Baja California (Fig. 3d). In addition, the typical low-frequency midlatitude pattern emanating from tropical convection associated with a strong El Niño was absent. In its place was an active waveguide with short-wavelength, high-amplitude troughs and ridges around the midlatitude Northern Hemisphere, including a period of record-level wave activity flux. A product of the active waveguide was a record-breaking trough that impacted Mexico with a wide variety of detrimental societal, economic, and environmental impacts. This study demonstrates the important role of the nature of the jet streams over the Pacific in regulating the response of the atmosphere to tropical convection. In this case, the extreme event over Mexico was not associated with a typical cold air outbreak moving from Canada to Mexico along the lee of the Rockies, as in past studies (e.g., Colle and Mass 1995; Magaña et al. 2003; Henry 1979; Reding 1992; Schultz et al. 1997; Pérez et al. 2014), but instead resulted from an interaction between tropical convection and a midlatitude RWP both injecting Rossby wave activity into the subtropical jet.

We can gain insight into the influence of the overall structure of the atmosphere as a refractive medium for Rossby wave activity, including fast narrow jet streams acting as waveguides. El Niño tropical convection did not play its typical role in California during this period, extending the historic drought. The combination of differences in tropical convection between El Niño years, and the differences in the nature of the waveguides, means a quite different response can occur over North America. However, the situation is not so simple as differences in seasonal and subseasonal tropical convection affecting the nature of the jet streams. The structure of the entire environmental PV field will be molded by other external conditions like subtropical and midlatitude SSTs, winds in the stratosphere, and internal dynamics in the angular momentum budget such as mountain and friction torque and eddy fluxes (e.g., Li and Wu 2010). In the case of the 2015/16 winter, we identify the unusually warm water outside of the ENSO region as one possible factor impacting the structure of PV and path of Rossby wave activity, persuading observed sensible weather to diverge from past strong El Niños, consistent with Quan et al. (2018). Recent studies utilizing ensembles of climate model simulations by Deser et al. (2017, 2018) demonstrate the variability in sensible weather that is possible in ENSO events. Our study reinforces this idea and elaborates on the pivotal role of the shape and characteristics of the waveguides in the Pacific in determining the downstream response to tropical convection over North America.

With the presence of a well-defined waveguide, the antecedent condition of long-lived RWPs coincided with increased forecast skill, consistent with studies on relationships between RWPs and intrinsic predictability (Grazzini and Vitart 2015). The RWP also acted to amplify and shorten low-frequency waves, which are typically easier to predict given persistence. Therefore, proper handling of these features allowed for the Mexico trough to be captured by medium-range NWP out to 10 days, along with signals at even longer lead time within subseasonal ensembles. While the existence of the RWP was properly modeled, NWP struggled to capture the strength of the RWP. The amplitude of the trough is shown to be dependent on the strength of the RWP itself. Ramifications of these discrepancies go beyond just the amplitude of the waves, including also wavelength, phase speed, and the timing of wave breaking. Consistent with this idea was maximum ensemble spread around wave nodes, indicating phase uncertainty. The spread associated with phase uncertainty is amplified by the RWP, making the wave packet a potential forecast skill liability.

The historic winter storm that impacted Mexico in March 2016 demonstrates the importance of our understanding of Rossby wave dynamics across all scales, and acknowledgment of variability beyond the canonical pictures of ENSO. This event, and the unique El Niño winter that facilitated it, should be a highlighted case in the endeavor to both 1) extend our limits of predictability for extreme high-impact weather and 2) improve our skill at subseasonal and seasonal time scales.

ACKNOWLEDGMENTS

The authors thank Steven Cavallo (OU), Nicholas Szapiro (Norwegian Meteorological Institute), Clara Deser (NCAR), and George Kiladis (NOAA/PSD) for their encouragement and feedback. We greatly appreciate the constructive comments from the three anonymous reviewers and from the editor that shaped the final form of this paper. This work was primarily supported using funds from the School of Meteorology’s Mark and Kandi McCasland endowed chair.

REFERENCES

REFERENCES
Bell
,
G. D.
, and
M. S.
Halpert
,
1998
:
Climate Assessment for 1997
.
Bull. Amer. Meteor. Soc.
,
79
(
Suppl
.), https://doi.org/10.1175/1520-0477-79.5s.S1.
Blunden
,
J.
, and
D. S.
Arndt
,
2016
:
State of the Climate in 2015
.
Bull. Amer. Meteor. Soc.
,
97
(
Suppl
.), https://doi.org/10.1175/2016BAMSStateoftheClimate.1.
Breaker
,
B. K.
,
K. M.
Watson
,
P. A.
Ensminger
,
J. B.
Storm
, and
C. E.
Rose
,
2016
: Characterization of peak streamflows and flood inundation of selected areas in Louisiana, Texas, Arkansas, and Mississippi from flood of March 2016. U.S. Geological Survey Tech. Rep., 43 pp.
Brower
,
L. P.
,
E. H.
Williams
,
P.
Jaramillo-López
,
D. R.
Kust
,
D. A.
Slayback
, and
M. I.
Ramírez
,
2017
:
Butterfly mortality and salvage logging from the March 2016 storm in the monarch butterfly biosphere reserve in Mexico
.
Amer. Entomol.
,
63
,
151
164
, https://doi.org/10.1093/ae/tmx052.
Caballero
,
R.
, and
B. T.
Anderson
,
2009
:
Impact of midlatitude stationary waves on regional Hadley cells and ENSO
.
Geophys. Res. Lett.
,
36
,
L17704
, https://doi.org/10.1029/2009GL039668.
Colle
,
B. A.
, and
C. F.
Mass
,
1995
:
The structure and evolution of cold surges east of the Rocky Mountains
.
Mon. Wea. Rev.
,
123
,
2577
2610
, https://doi.org/10.1175/1520-0493(1995)123<2577:TSAEOC>2.0.CO;2.
Cullather
,
R. I.
,
Y.-K.
Lim
,
L. N.
Boisvert
,
L.
Brucker
,
J. N.
Lee
, and
S. M.
Nowicki
,
2016
:
Analysis of the warmest Arctic winter
,
2015–2016
.
Geophys. Res. Lett.
,
43
,
10 808
10 816
, https://doi.org/10.1002/2016GL071228.
Dee
,
D. P.
, and Coauthors
,
2011
:
The ERA-Interim reanalysis: Configuration and performance of the data assimilation system
.
Quart. J. Roy. Meteor. Soc.
,
137
,
553
597
, https://doi.org/10.1002/qj.828.
Deser
,
C.
,
I. R.
Simpson
,
K. A.
McKinnon
, and
A. S.
Phillips
,
2017
:
The Northern Hemisphere extratropical atmospheric circulation response to ENSO: How well do we know it and how do we evaluate models accordingly?
J. Climate
,
30
,
5059
5082
, https://doi.org/10.1175/JCLI-D-16-0844.1.
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
,
4991
5014
, https://doi.org/10.1175/JCLI-D-17-0783.1.
Dole
,
R. M.
, and Coauthors
,
2017
:
Advancing science and services during the 2015/16 El Niño: The NOAA El Niño Rapid Response field campaign
.
Bull. Amer. Meteor. Soc.
,
99
,
975
1001
, https://doi.org/10.1175/BAMS-D-16-0219.1.
Grazzini
,
F.
, and
F.
Vitart
,
2015
:
Atmospheric predictability and Rossby wave packets
.
Quart. J. Roy. Meteor. Soc.
,
141
,
2793
2802
, https://doi.org/10.1002/qj.2564.
Held
,
I. M.
,
1983
: Stationary and quasi-stationary eddies in the extratropical troposphere: Theory. Large-Scale Dynamical Processes in the Atmosphere, B. Hoskins and R. Pearce, Eds., Academic Press, 127–168.
Held
,
I. M.
, and
A. Y.
Hou
,
1980
:
Nonlinear axially symmetric circulations in a nearly inviscid atmosphere
.
J. Atmos. Sci.
,
37
,
515
533
, https://doi.org/10.1175/1520-0469(1980)037<0515:NASCIA>2.0.CO;2.
Henry
,
W.
,
1979
:
Some aspects of the fate of cold fronts in the Gulf of Mexico
.
Mon. Wea. Rev.
,
107
,
1078
1082
, https://doi.org/10.1175/1520-0493(1979)107<1078:SAOTFO>2.0.CO;2.
Hoskins
,
B. J.
, and
D. J.
Karoly
,
1981
:
The steady linear response of a spherical atmosphere to thermal and orographic forcing
.
J. Atmos. Sci.
,
38
,
1179
1196
, https://doi.org/10.1175/1520-0469(1981)038<1179:TSLROA>2.0.CO;2.
Hoskins
,
B. J.
, and
T.
Ambrizzi
,
1993
:
Rossby wave propagation on a realistic longitudinally varying flow
.
J. Atmos. Sci.
,
50
,
1661
1671
, https://doi.org/10.1175/1520-0469 (1993)050<1661:RWPOAR>2.0.CO;2.
Jong
,
B.-T.
,
M.
Ting
, and
R.
Seager
,
2016
:
El Niño’s impact on California precipitation: Seasonality, regionality, and El Niño intensity
.
Environ. Res. Lett.
,
11
, 054021, https://doi.org/10.1088/1748-9326/11/5/054021.
L’Heureux
,
M. L.
, and Coauthors
,
2017
:
Observing and predicting the 2015/16 El Niño
.
Bull. Amer. Meteor. Soc.
,
98
,
1363
1382
, https://doi.org/10.1175/BAMS-D-16-0009.1.
Li
,
J.
, and
G.
Wu
,
2010
:
Atmospheric angular momentum transport and balance in the AGCM-SAMIL
.
Adv. Atmos. Sci.
,
27
,
1183
1192
, https://doi.org/10.1007/s00376-009-9157-5.
Luna-Niño
,
R.
, and
T.
Cavazos
,
2018
:
Formation of a coastal barrier jet in the Gulf of Mexico due to the interaction of cold fronts with the Sierra Madre oriental mountain range
.
Quart. J. Roy. Meteor. Soc.
,
144
,
115
128
, https://doi.org/10.1002/qj.3188.
Magaña
,
V.
, and
T.
Ambrizzi
,
2005
:
Dynamics of subtropical vertical motions over the Americas during El Niño boreal winters
.
Atmósfera
,
18
,
211
235
.
Magaña
,
V.
,
J. L.
Vázquez
,
J. L.
Pérez
, and
J. B.
Pérez
,
2003
:
Impact of El Niño on precipitation in Mexico
.
Geofis. Int.
,
42
,
313
330
.
Matsuno
,
T.
,
1970
:
Vertical propagation of stationary planetary waves in the winter Northern Hemisphere
.
J. Atmos. Sci.
,
27
,
871
883
, https://doi.org/10.1175/1520-0469(1970)027<0871:VPOSPW>2.0.CO;2.
Mesinger
,
F.
, and Coauthors
,
2006
:
North American Regional Reanalysis
.
Bull. Amer. Meteor. Soc.
,
87
,
343
360
, https://doi.org/10.1175/BAMS-87-3-343.
Nakamura
,
H.
,
1992
:
Midwinter suppression of baroclinic wave activity in the Pacific
.
J. Atmos. Sci.
,
49
,
1629
1642
, https://doi.org/10.1175/1520-0469 (1992)049<1629:MSOBWA>2.0.CO;2.
Newman
,
M.
,
G. P.
Compo
, and
M. A.
Alexander
,
2003
:
ENSO-forced variability of the Pacific decadal oscillation
.
J. Climate
,
16
,
3853
3857
, https://doi.org/10.1175/1520-0442(2003)016<3853:EVOTPD>2.0.CO;2.
Oort
,
A. H.
, and
J. J.
Yienger
,
1996
:
Observed interannual variability in the Hadley circulation and its connection to ENSO
.
J. Climate
,
9
,
2751
2767
, https://doi.org/10.1175/1520-0442(1996)009<2751:OIVITH>2.0.CO;2.
O’Rourke
,
A. K.
, and
G. K.
Vallis
,
2016
:
Meridional Rossby wave generation and propagation in the maintenance of the wintertime tropospheric double jet
.
J. Atmos. Sci.
,
73
,
2179
2201
, https://doi.org/10.1175/JAS-D-15-0197.1.
Pérez
,
E. P.
,
V.
Magaña
,
E.
Caetano
, and
S.
Kusunoki
,
2014
:
Cold surge activity over the Gulf of Mexico in a warmer climate
.
Front. Earth Sci.
,
2
,
19
, https://doi.org/10.3389/feart.2014.00019.
Peterson
,
W.
,
M.
Robert
, and
N.
Bond
,
2015
: The warm blob continues to dominate the ecosystem of the Northern California current. PICES Press, Vol. 23, No. 2, North Pacific Marine Science Organization, Sidney, BC, Canada, 44–46.
Quan
,
X.-W.
,
M.
Hoerling
,
L.
Smith
,
J.
Perlwitz
,
T.
Zhang
,
A.
Hoell
,
K.
Wolter
, and
J.
Eischeid
,
2018
:
Extreme California rains during winter 2015/16: A change in El Niño teleconnection [in “Explaining Extreme Events of 2016 from a Climate Perspective”]?
Bull. Amer. Meteor. Soc.
,
99
(
1
),
S49
S53
, https://doi.org/10.1175/BAMS-D-17-0118.1.
Reding
,
P. J.
,
1992
: The Central American cold surge: An observational analysis of the deep southward penetration of North American cold fronts. M.S. thesis, Dept. of Meteorology, Texas A&M University, 177 pp.
Ropelewski
,
C. F.
, and
M. S.
Halpert
,
1987
:
Global and regional scale precipitation patterns associated with the El Niño/Southern Oscillation
.
Mon. Wea. Rev.
,
115
,
1606
1626
, https://doi.org/10.1175/1520-0493(1987)115<1606:GARSPP>2.0.CO;2.
Rosen
,
R. D.
,
D. A.
Salstein
,
T. M.
Eubanks
,
J. O.
Dickey
, and
J. A.
Steppe
,
1984
:
An El Nino signal in atmospheric angular momentum and Earth rotation
.
Science
,
225
,
411
414
, https://doi.org/10.1126/science.225.4660.411.
Santoso
,
A.
,
M. J.
Mcphaden
, and
W.
Cai
,
2017
:
The defining characteristics of ENSO extremes and the strong 2015/2016 El Niño
.
Rev. Geophys.
,
55
,
1079
1129
, https://doi.org/10.1002/2017RG000560.
Sardeshmukh
,
P. D.
, and
B. J.
Hoskins
,
1988
:
The generation of global rotational flow by steady idealized tropical divergence
.
J. Atmos. Sci.
,
45
,
1228
1251
, https://doi.org/10.1175/1520-0469(1988)045<1228:TGOGRF>2.0.CO;2.
Schaefer
,
J.
, and
R.
Edwards
,
1999
: The SPC tornado/severe thunderstorm database. Preprints, 11th Conf. on Applied Climatology, Dallas, TX, Amer. Meteor. Soc., 603–606.
Schneider
,
E. K.
,
1987
:
A simplified model of the modified Hadley circulation
.
J. Atmos. Sci.
,
44
,
3311
3328
, https://doi.org/10.1175/1520-0469(1987)044<3311:ASMOTM>2.0.CO;2.
Schultz
,
D. M.
,
W. E.
Bracken
,
L. F.
Bosart
,
G. J.
Hakim
,
M. A.
Bedrick
,
M. J.
Dickinson
, and
K. R.
Tyle
,
1997
:
The 1993 superstorm cold surge: Frontal structure, gap flow, and tropical impact
.
Mon. Wea. Rev.
,
125
,
5
39
, https://doi.org/10.1175/1520-0493(1997)125<0005:TSCSFS>2.0.CO;2.
Schultz
,
D. M.
,
W. E.
Bracken
, and
L. F.
Bosart
,
1998
:
Planetary- and synoptic-scale signatures associated with Central American cold surges
.
Mon. Wea. Rev.
,
126
,
5
27
, https://doi.org/10.1175/1520-0493(1998)126<0005:PASSSA>2.0.CO;2.
Shapiro
,
M. A.
, and
A. J.
Thorpe
,
2004
: Thorpex international science plan. WMO Rep. WMO/TD-1246, 57 pp.
Shapiro
,
M. A.
, and Coauthors
,
2010
:
An Earth-system prediction initiative for the twenty-first century
.
Bull. Amer. Meteor. Soc.
,
91
,
1377
1388
, https://doi.org/10.1175/2010BAMS2944.1.
Stein
,
A.
,
R. R.
Draxler
,
G. D.
Rolph
,
B. J.
Stunder
,
M.
Cohen
, and
F.
Ngan
,
2015
:
NOAA’s HYSPLIT atmospheric transport and dispersion modeling system
.
Bull. Amer. Meteor. Soc.
,
96
,
2059
2077
, https://doi.org/10.1175/BAMS-D-14-00110.1.
Takaya
,
K.
, and
H.
Nakamura
,
1997
:
A formulation of a wave-activity flux for stationary Rossby waves on a zonally varying basic flow
.
Geophys. Res. Lett.
,
24
,
2985
2988
, https://doi.org/10.1029/97GL03094.
Takaya
,
K.
, and
H.
Nakamura
,
2001
:
A formulation of a phase-independent wave-activity flux for stationary and migratory quasigeostrophic eddies on a zonally varying basic flow
.
J. Atmos. Sci.
,
58
,
608
627
, https://doi.org/10.1175/1520-0469(2001)058<0608:AFOAPI>2.0.CO;2.
Teng
,
H.
, and
G.
Branstator
,
2017
:
Causes of extreme ridges that induce California droughts
.
J. Climate
,
30
,
1477
1492
, https://doi.org/10.1175/JCLI-D-16-0524.1.
Thorpe
,
A. J.
,
2004
:
Weather forecasting: A centenary perspective
.
Weather
,
59
,
332
335
, https://doi.org/10.1256/wea.87.04.
Uccellini
,
L. W.
,
2016
: Service assessment: The historic nor’easter of January 2016. National Weather Service Rep., 83 pp.
Wallace
,
J. M.
, and
D. S.
Gutzler
,
1981
:
Teleconnections in the geopotential height field during the Northern Hemisphere winter
.
Mon. Wea. Rev.
,
109
,
784
812
, https://doi.org/10.1175/1520-0493(1981)109<0784:TITGHF>2.0.CO;2.
Wirth
,
V.
, and
J.
Eichhorn
,
2014
:
Long-lived Rossby wave trains as precursors to strong winter cyclones over Europe
.
Quart. J. Roy. Meteor. Soc.
,
140
,
729
737
, https://doi.org/10.1002/qj.2191.
Wirth
,
V.
,
M.
Riemer
,
E. K.
Chang
, and
O.
Martius
,
2018
:
Rossby wave packets on the midlatitude waveguide—A review
.
Mon. Wea. Rev.
,
146
,
1965
2001
, https://doi.org/10.1175/MWR-D-16-0483.1.
Wolf
,
G.
, and
V.
Wirth
,
2017
:
Diagnosing the horizontal propagation of Rossby wave packets along the midlatitude waveguide
.
Mon. Wea. Rev.
,
145
,
3247
3264
, https://doi.org/10.1175/MWR-D-16-0355.1.

Footnotes

© 2019 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).