1. Introduction
Cool-season long-lived and widespread extreme precipitation events often lead to substantial societal and economic impacts, including the loss of life, property, and infrastructure. One example is the multiple storms composing the historic ice storm of January 1998, which affected parts of northern New York (NY), northern New England, eastern Ontario, and southern Quebec with up to 100 mm of ice accumulation (along with flooding farther south), leading to millions of power outages, 56 fatalities (including those resulting from flooding), and billions of U.S. dollars of damages (e.g., Gyakum and Roebber 2001). Another example is the persistent cyclone track from the southern Rockies/southern Plains region to New England during the winter of 2007–08 that led to an unusually wet winter in a large section of the eastern United States and a record snowy winter in parts of southern Ontario and southern Quebec, resulting in near-record floods, large snow removal costs, and the collapse of several buildings (Peterson and Baringer 2009; Changnon et al. 2008).
Quantitative precipitation forecasting (QPF) is still a formidable challenge in operational meteorology today, especially for extreme precipitation events. Although the accuracy of model simulations of extratropical cyclones and other synoptic-scale features responsible for the precipitation has improved substantially over the past few decades, improvements in QPF have been slower (Milrad et al. 2009; Sharma et al. 2017). Therefore, it is important to identify particular patterns and/or precursors associated with such events to aid with forecasting of such events (Milrad et al. 2009). Pattern recognition and knowledge has been demonstrated to improve on pure model QPF, even for lead times as short as 2 or 3 days. This is especially true for unusual or extreme events, when the forecaster can use pattern recognition to deviate substantially from model statistical guidance for the better (Novak et al. 2014). A more complete understanding of these events, especially longer-lived events, could also play an important role in improving the prediction of such events at subseasonal to seasonal (S2S) lead times (Jennrich et al. 2020).
In midlatitudes, heavy precipitation is usually produced by mobile, transient weather systems (e.g., extratropical cyclones; Pfahl and Wernli 2012; Wallace et al. 1988) and is therefore short-lived in a particular region. As such, most previous studies on extreme precipitation over North America have focused on a single event (e.g., Bosart 1981; Lackmann and Gyakum 1996; Lackmann et al. 1998; Lackmann and Gyakum 1999; Gyakum and Roebber 2001; Milrad et al. 2015), a collection of individual storms for compositing (e.g., Lackmann and Gyakum 1996, 1999; Milrad et al. 2009, 2010b), or events of duration no longer than 5 days (e.g., Archambault et al. 2008; Gyakum 2008; Archambault et al. 2010; Milrad et al. 2010a; Moore et al. 2015, 2019). These studies found teleconnections, anomalous large-scale flow patterns, and/or enhanced poleward moisture transport that result in short-duration extreme precipitation events. Similarly, a number of other studies have also shown that atmospheric rivers (ARs) or related features (e.g., warm conveyor belts and tropical moisture exports) are major contributors to at least short-lived precipitation events and are responsible for a substantial fraction of total seasonal and/or annual precipitation in eastern North America (e.g., Eckhardt et al. 2004; Knippertz and Wernli 2010; Lavers and Villarini 2013; Pfahl et al. 2014; Lavers and Villarini 2015; Mahoney et al. 2016). More persistent heavy precipitation is favored when an AR stalls (e.g., Ralph et al. 2019) or when cyclones cluster such that successive ARs affect the same region (Fish et al. 2019).
Heavy precipitation can be long-lived if a low pressure system stalls and becomes quasi-stationary (e.g., Lenggenhager et al. 2019; Doswell et al. 1998). Alternatively, multiple low pressure systems can track over the same region; this process is often crucial for heavy precipitation events to be prolonged (e.g., Lackmann and Gyakum 1999; Sodemann and Stohl 2013; Priestley et al. 2017b; Moore et al. 2020). Studies of these clustered systems will provide insight into long-duration precipitation events in eastern North America, which is affected by three major cyclone tracks: those of Alberta Clippers, Colorado cyclones, and East Coast cyclones. These cyclone tracks differ in evolution, quasigeostrophic (QG) dynamics, stability profiles, and moisture profiles, with the Colorado and East Coast cyclone tracks being associated with the strongest dynamics and most moisture (Mercer and Richman 2007), implying the heaviest precipitation. Sodemann and Stohl (2013) found that the anomalous warmth and heavy precipitation in Norway in December 2006 were associated with warm conveyor belts of multiple midlatitude cyclones feeding off ARs (through large-scale ascent and precipitation) and reinforcing ARs. Pinto et al. (2014), Priestley et al. (2017b), and Priestley et al. (2017a) found that wintertime cyclone clustering in western Europe is associated with a strong, straight, and persistent North Atlantic jet with a recurrent extension into western Europe and Rossby wave breaking on both flanks of the jet. Mailier et al. (2006) suggests that the time-varying effect of the large-scale flow on individual cyclone tracks and the generation by one “parent” cyclone of one or more “offspring” cyclones through secondary cyclogenesis are possible mechanisms for repeated cyclones.
A few studies have examined longer-duration precipitation periods on a systematic climatological basis. Moore et al. (2021) found that long-duration (≥3 days) heavy precipitation events along the U.S. West Coast are characterized by a strong zonal jet stream over the eastern North Pacific or atmospheric blocking over the central North Pacific or the Bering Sea–Alaska regions. The flow patterns tend to remain in place for several days, maintaining strong baroclinicity and promoting a rapid succession of multiple cyclones near the West Coast. Jennrich et al. (2020) identified 14-day extreme precipitation events in the contiguous United States partitioned into six different geographic regions. They found common synoptic signals during events such as zonally oriented 500-hPa trough–ridge patterns, a strong subtropical jet stream, and enhanced moisture transport into the affected region. They also found increases in AR activity in the affected region, synoptic-scale wind and geopotential height anomalies in the North Pacific, and regional links to teleconnections.
Our study aims to analyze extreme precipitation regimes (EPRs) in a new, more comprehensive manner by expanding the work on precipitation events to include extreme, long-duration, and large-scale precipitation periods during the eastern North American winter. We define EPRs as persistent periods of extreme precipitation volume, normalized by climatology, and analyze them during the eastern North American winter. High-impact weather events associated with EPRs include, but are not limited to, the 1998 ice storm and flooding; the wet and snowy winter of 2007–08; and major flooding in the midwestern and southern United States during January–February 1982 (Changnon et al. 1983; Wagner 1982), December 1982–January 1983 (Sauer and Fulford 1983), December 2015 (NOAA/NCEI 2016), February 2018 (NOAA/NCEI 2018), and February 2019 (NOAA/NCEI 2019). Our long-duration threshold is distinct from the previous studies of individual storms or short-duration events listed above, and our region of study is distinct from the U.S. West Coast region that Moore et al. (2021) studied. Unlike in Jennrich et al. (2020), we do not split eastern North America into subregions, our EPRs can be of variable durations, and our thresholds to define extreme and widespread precipitation are based on seasonally varying climatology rather than fixed thresholds throughout the year. We aim to analyze the synoptic-scale thermodynamic and dynamic structures during the EPRs in substantial detail. We also compare and contrast our results with those in past studies.
Section 2 describes the dataset, region, and methodology used, including the EPR definition, the synoptic-scale and thermodynamic metrics used for analyzing EPRs, and the EPR categorization scheme. Section 3 describes the spatial and temporal distribution of EPRs. Section 4 describes the composite analysis of the metrics during EPRs. Section 5 describes a comparison between the two EPR categories. Section 6 highlights some key conclusions of the analysis and describes a comparison with previous studies on precipitation periods.
2. Methodology
a. Defining EPRs
To define and analyze EPRs, we use the precipitation variable from the ERA5 dataset, which has a 0.25° × 0.25° spatial resolution and a 1-h temporal resolution, and is openly available through the Copernicus Climate Change Service (Hersbach et al. 2020). Although like with many reanalyses, the ERA5 precipitation may not be accurate with convective precipitation; it is otherwise accurate (Beck et al. 2019; Crossett et al. 2020; Tarek et al. 2020), and therefore presumably satisfactory during winter on the large-scale. In addition, the precipitation normalization scheme described later reduces the contribution of convective precipitation in the southern United States. We also computed the EPR dates with the NARR, which despite having precipitation issues over Canada (Milrad et al. 2013) that are not present in the ERA5, yields very similar EPRs, so the EPR dates are not sensitive to the precipitation dataset used. Also, since we use ERA5 for the rest of the analysis, using ERA5 would make all the data consistent.
In this study, we focus on eastern North America, which we define as the region of 25°–50°N and 95°–50°W that is over land (the region in the black box and shaded in brown in Fig. 1a), and includes the heavily populated regions of Canada, from Toronto to Quebec City, and of the United States, from Washington, D.C., to Boston, Massachusetts.
(a) Map of eastern North America, depicted by the brown shading in the black box. Abbreviations or names of the U.S. states and Canadian provinces mentioned in the text are annotated. (b) Map of the simplified eastern North American region, depicted by the brown shading, used to calculate the moisture budget. The blue, magenta, cyan, and yellow lines depict the northern, western, Gulf of Mexico, and East Coast boundary segments, which are denoted by “north,” “west,” “GofM,” and “EC,” respectively. (c) As in (a), but for the domain used to calculate the anomaly correlations (ACs) described later.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
Illustration of the EPR definition. (a) Winter 80th percentile of measurable daily precipitation amounts (p80th). (b) Illustration of the extreme threshold, showing the actual precipitation volume (green; mm km2) from 17 to 27 Feb 2018 and the 70th percentile of precipitation volume (the extreme threshold; blue; mm km2). The extreme threshold is satisfied for 7 consecutive days (19–25 Feb, shown as the black interval), making this period an EPR. (c) 1-day precipitation on 20 Feb 2018; (d) 1-day normalized precipitation on 20 Feb 2018; (e),(f) As in (c) and (d), but for 3 Dec 1983.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
Next, we calculate the 70th percentile of Vnorm over all winter days, which is the extreme threshold for the whole eastern North American domain. This threshold is strict enough to eliminate most non-impactful precipitation events but lenient enough to produce a sufficient sample size for a composite analysis. Finally, an EPR occurs when the extreme threshold is exceeded for at least 5 consecutive days, except for breaks of at most 1 day, as long as at least two-thirds of days meet the threshold. This definition allows for impactful heavy precipitation events with brief breaks to still be counted. We show an EPR example in Fig. 2b.
The EPR definition contrasts with that of Jennrich et al. (2020) and Moore et al. (2021) and allows us to detect events that are noteworthy for winter but not necessarily for other seasons and to weigh different parts of the region based on its precipitation climatology. It excludes high impact events that are short-duration, localized, or are longer duration but have large gaps, such as the 5–10 February 2010 period, which featured two historic snowstorms in the mid-Atlantic United States but a 3-day gap in between the two storms.
b. Calculating EPR centroids
c. Categorizing EPRs
(a) Average total precipitation (shaded according to color bar; mm) and its percentile (blue contours at intervals of 5 at and above 70) during EPRs. Gray dots mark individual EPR centroids and the purple dot marks the composite EPR centroid. The gray boundary encloses the eastern North American region. (b) Histogram of number of EPRs vs EPR duration. (c) Scattergram of EPR region-averaged precipitation rate (mm day−1) vs EPR duration (days). The solid blue line represents the best linear fit between EPR region-averaged precipitation rate and EPR duration, and the dashed blue line represents the climatological winter region-averaged precipitation rate. (d) Scattergram of EPR 1000–500-hPa AC vs SLP AC. The red horizontal and vertical lines represent the two-thirds threshold for SLP and 1000–500-hPa ACs, respectively. Circles and squares indicate SLP ACs of below and above the two-thirds threshold, respectively. Blue and green markers indicate 1000–500-hPa ACs of below and above the two-thirds threshold, respectively. The black rectangle marks the SLP and 1000–500-hPa AC space occupied by high AC EPRs.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
d. Analyzing EPRs through composites
1) Precipitation
First, we composite the total precipitation amount and the precipitation rate percentile during EPRs. At a given grid point, the precipitation rate percentile of an EPR of length n days is the percentile of all winter periods of length n with measurable precipitation corresponding to the average precipitation rate during the EPR.
2) Synoptic-scale and thermodynamic patterns
3) 500-hPa variability
4) Cyclone and anticyclone density
Next, using the ERA5 SLP variable, we compute surface cyclone and anticyclone frequencies during all winters, representing the climatology, and during EPRs. We compute the EPR frequency anomalies with respect to the climatological frequencies. To compute the frequencies, we first define a low (high) pressure center at a given time to be a minimum (maximum) in SLP that is at most 1015 hPa (at least 1020 hPa) and at least 0.5 (0.2) hPa lower (higher) than at either both 150 km to the north and south or both 150 km to the west and east. Then, we pair each center with a low (high) pressure center found in the following 4 h and within 500 km of the center, if there is one, and connect the two centers to form a cyclone (anticyclone) track. We repeat this step until there is no suitable pairing, signaling the end of the cyclone (anticyclone) track. We only include centers that persist for at least 6 h. For the EPR cyclone and anticyclone frequencies, we count the persistent centers that are present for any nonzero duration during the EPR. These criteria filter out low (high) pressure systems that are very weak, broad, elongated, and/or short-lived while keeping those that are relevant for at least part of the EPR’s lifetime. Next, we define a low (high) pressure footprint to be the circle of radius 400 km from each point of the low (high) pressure center track. Finally, the average cyclone (anticyclone) frequency is the frequency of being in a low (high) pressure footprint. This metric takes into account both the number and persistence or lack of movement of cyclones and anticyclones over a region. The detection and tracking algorithm and thresholds are similar to but more inclusive than those in Wang et al. (2006) in order to detect systems with a weaker or elongated pressure gradient that could still be dynamically important and to account for possible jumps in low or high pressure centers that would artificially reduce the calculated longevity of the systems. The footprints are similar to those in Bentley et al. (2019). We obtain similar qualitative results by changing the SLP, SLP gradient, and center tracking thresholds modestly (not shown).
5) Moisture budget
6) ARs
Finally, we detect ARs using the Pan and Lu (2019) methodology and analyze AR frequencies of different AR categories in eastern North America during EPRs. Essentially, the Pan and Lu (2019) algorithm finds contiguous grid points exceeding dual IVT thresholds: a local threshold (85th percentile of IVT at a grid point) and a regional threshold (80th percentile of the IVT field over all grid points in a specified region). We detect an AR if the region composed of the grid points is ≥2000 km long and has a length to width ratio of ≥2. For the regional threshold, the specified region we use is the same latitude–longitude box that we use to define eastern North America, but including the ocean regions, since ARs frequently originate and/or traverse over oceans. We exclude the spatial smoothing done in Pan and Lu (2019) to calculate the local threshold for simplicity, given that the IVT field over eastern North America usually does not have sharp gradients. We also exclude the maximum width criterion in Pan and Lu (2019), as we are also interested in wide regions of moisture transport. As discussed later, the results are not sensitive to the exact AR definition.
7) Statistical significance
We determine the statistical significance of the 500-hPa height, θe,max, and cyclone and anticyclone density anomalies based on the field significance approach described in Wilks (2016) using a test level of αfalse-discovery-rate = 2αglobal suggested for well-correlated spatial fields, and choosing αglobal = 0.05. For this determination, we coarsen the 500-hPa height anomaly field to 2° × 2° and the other fields to 1° × 1° to reduce interdependency between grid points. There is sufficient temporal separation between most EPRs such that they are independent of each other. We base the local hypothesis tests on Student’s t tests with the null hypothesis being the EPR mean anomaly being equal to the climatological anomaly, which is zero by definition.
3. Temporal and spatial distribution
Using the methodology above and the ERA5 dataset, we find 62 EPRs, including 29 low AC EPRs and 33 high AC EPRs, in eastern North America during the winters (December–February) from 1979–80 to 2019–20. Figure 3 depicts the temporal and spatial distribution of EPRs, as well as the relationship between region-averaged precipitation rate and EPR duration. Figure 3a shows that EPRs impact a large proportion of eastern North America with heavy precipitation, though precipitation is notably less over Minnesota (MN), northern Ontario, and the Florida (FL) peninsula, which correspond to the northwestern and southeastern corners of eastern North America. The greatest precipitation occurs in a broad southwest–northeast-oriented region from Louisiana (LA) to Quebec, including the Deep South, Ohio Valley, mid-Atlantic, and Northeast regions of the United States, as well as southern Ontario, southern Quebec, and the Canadian Maritimes. These regions experience >70th percentile, with some areas experiencing >80th percentile, of precipitation total for the EPR duration on average. The composite EPR centroid is located in southwestern Ontario, with all of the individual centroids being located within ±3° latitude and 5° longitude. We hypothesize that the relative lack of EPR centroid variability could be due to the widespread nature of precipitation during EPRs. The EPR criteria are difficult to satisfy with only localized areas of precipitation.
Figure 3b shows that the number of EPRs decreases rapidly with increasing duration; 36 of the 62 EPRs have a duration of 5 days, the shortest possible, and the vast majority of EPRs (57 of 62) have a duration between 5 and 8 days. EPRs are more common in December than in January and February, with 28, 13, and 21 EPRs having their midtimes in December, January, and February, respectively. This pattern could be due to the water temperature being warmer and able to provide more moisture early in the winter, as well as our usage of one value for the 80th percentile of daily precipitation for the entire winter. There is no clear correlation between EPR region-averaged precipitation rate and EPR duration, as shown by the linear fit in Fig. 3c. The lack of a clear correlation could be due to substantial variability between precipitation rates within EPRs of the same duration and/or similar dynamics and thermodynamics regardless of EPR duration, except for the persistence; this finding is also supported by the composites of EPRs of different durations (not shown). Despite the variability, all EPRs have an above climatological region-averaged precipitation rate (Fig. 3c). Figure 3d shows that there is a substantially greater variability in both SLP and 1000–500-hPa thickness AC values, indicating more variable weather patterns, among low AC EPRs, than among high AC EPRs.
4. Composite analysis
a. 500-hPa pattern evolution
Figure 4 depicts the composite 500-hPa geopotential height pattern prior to and during EPRs. Three days before the EPR start (ts−3, Fig. 4a), no substantial signals are found in North America. Two days before the EPR start (ts−2, Fig. 4b), an anomalous trough is present in western North America and a slight anomalous ridge is present in the eastern United States. One day before the EPR start, (ts−1, Fig. 4c), the anomalous trough persists while the anomalous ridge strengthens and expands. From ts−1, the pattern moves slightly eastward through the EPR midtime (tm), with the ridge being slightly stronger at the EPR start (ts, Fig. 4d) and tm (Fig. 4e). By the EPR end (te, Fig. 4f), the anomalous trough–ridge pattern in North America has moved eastward substantially. During the EPR’s lifetime, the eastern U.S. ridge is more anomalous than the western North America trough. In addition, a near-stationary anomalous ridge is present in the central North Pacific throughout the EPR’s lifetime. This anomalous trough–ridge pattern is also quite robust throughout the EPR’s lifetime (shown in stippling), especially at ts and tm, indicating the persistent nature of the pattern.
(a) Composite 500-hPa height (solid contours, 6 dam interval) and anomaly (shaded according to color bar; dam) for 3 days before EPR start (ts−3). (b) As in (a), but for 2 days before EPR start (ts−2). (c) As in (a), but for 1 day before EPR start (ts−1). (d) As in (a), but for EPR start (ts). (e) As in (a), but for EPR midtime (tm). (f) As in (a), but for EPR end (te). Stippling indicates field statistical significance of anomalies at the αfalse-discovery-rate = 0.1 or αglobal = 0.05 level. The yellow boundary encloses the eastern North American region, and the navy blue dot marks the composite EPR centroid.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
b. Thermodynamics and precipitation evolution
Figure 5 depicts the θe,max anomaly, wind anomalies at the level of θe,max, SLP, and precipitation patterns at ts, tm, and te. It shows anomalously and robustly high θe,max over almost all of eastern North America and an anticyclone in the subtropical western Atlantic, associated with the eastern U.S. 500-hPa ridge, throughout the EPR’s lifetimes. This pattern is associated with anomalous south to southwesterly flow from the Gulf of Mexico both at the surface (shown by the SLP patterns) and at the level of θe,max (shown by the vectors) transporting warmth and moisture northward. There is also an anomalously strong θe,max gradient from the central United States to Atlantic Canada. However, the spatial patterns vary noticeably during the EPR’s lifetimes. At ts (Figs. 5a,b), the greatest θe,max anomalies are over the Great Lakes and southern Mississippi Valley, east of a diffuse low pressure region centered over Colorado (CO) and just northwest of the region of heaviest precipitation in the southern United States. At tm (Figs. 5c,d), the region of low pressure mostly disappears, but the high θe,max anomalies and heavy precipitation are still present though shifted and expanded eastward into Atlantic Canada. At te (Figs. 5e,f), a diffuse low pressure appears again but in Quebec, and the high θe,max anomalies and heavy precipitation are restricted offshore except for eastern New England and Atlantic Canada. Unlike earlier in the EPR, negative θe,max anomalies appear farther west over the midwestern and southern United States. These changes indicate that the warmth, moisture, and precipitation move offshore as the EPR ends.
(a) Composite θe,max anomaly (shaded according to color bar; K) with wind anomaly vectors (reference vector at top left of figure) at the level of θe,max and SLP (blue contours every 2 hPa) at ts. The green dot marks the composite EPR centroid. (b) Composite 24-h precipitation amount (shaded in mm according to color bar) centered at ts. The purple dot marks the composite EPR centroid. (c),(d) As in (a) and (b), but for tm. (e),(f) As in (a) and (b), but for te. Low and high pressure centers are labeled with red “L” and purple “H” markers, respectively. Stippling indicates field statistical significance of anomalies at the αfalse-discovery-rate = 0.1 or αglobal = 0.05 level. The gray boundary encloses the eastern North American region.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
Though the general patterns are robust, indicated by Fig. 3d, showing that most of the individual EPRs correlate very well with the EPR composite, the exact magnitude of the anomalies and the differences between ts, tm, and te are slightly sensitive to the thresholds used in the EPR definition. Also, since many EPRs are characterized by a series of surface cyclones, and the exact timing and location of the individual cyclones vary across EPRs (not shown), the composite SLP patterns are smeared out. We also find that in many EPRs, dry northerly flow behind each cyclone causes short breaks in precipitation, warmth, and moisture (not shown).
c. Surface low and high pressure center frequencies
Figure 6a depicts the EPR surface cyclone frequency anomalies. There is a region of anomalously high cyclone frequency from Texas (TX) northeastward into Michigan (MI) and a secondary region from Virginia (VA) to New Hampshire (NH), which appear to correspond to anomalously active Colorado low and Miller B cyclone tracks, respectively (Mercer and Richman 2007; Miller 1946), with the Colorado low track often shifted southward into Texas (TX). Many of the Miller B cyclones are continuations of the Colorado lows (not shown). The more active Colorado/southern plains low cyclone track during EPRs is favorable for transporting abundant moisture from the Gulf of Mexico (Mercer and Richman 2007) and heavy precipitation in eastern North America. These results are consistent with those of Bentley et al. (2019), who found that extreme weather events in eastern North America are much more likely to occur with Colorado lows, cyclones forming over the south-central United States, and East Coast cyclones tracking northeastward than with Alberta Clippers. There is also a region of anomalously low cyclone frequency offshore in the northwestern Atlantic, as offshore cyclones typically produce the majority of associated precipitation in the ocean, which is not included in our eastern North American region and therefore do not contribute substantially to EPRs.
(a) Cyclone frequency anomaly during EPRs shaded according to color bar, with EPR IVT anomaly vectors (reference vector on top right). Frequency represents the fraction of EPR’s lifetimes in a high or low pressure footprint. (b) As in (a), but for anticyclone frequency anomalies. The magenta dot marks the composite EPR centroid. Stippling indicates field statistical significance of anomalies at the αfalse-discovery-rate = 0.1 or αglobal = 0.05 level. The black boundary encloses the eastern North American region.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
Figure 6b depicts the EPR surface anticyclone frequency anomalies. There is a region of anomalously high anticyclone frequency from northern Quebec (QC) eastward to Labrador and a secondary region in the subtropical western Atlantic. The anticyclones to the north advect cold air southward and subtropical western Atlantic anticyclones advect warm air northward, increasing the baroclinicity. There is also an anomalous lack of precipitation-suppressing anticyclones over most of eastern North America itself. The mean IVT is southwesterly during both EPRs and climatology, but it is stronger with a greater southerly component during EPRs, which supports the anomalously large, AR-induced Gulf of Mexico flux [MGofM in Eq. (14)] shown later. The IVT rapidly weakens moving northward into the region of anomalously high cyclone frequency in the upper Midwestern United States. The greatest and most expansive positive anticyclone frequency anomalies occur at ts, in contrast to the cyclone frequency anomalies, which are negative throughout all of eastern North America at ts before becoming positive in most of eastern North America by tm and positive only closer to the East Coast by te (not shown). These results are consistent with the SLP composites in Fig. 5.
d. ARs
Figure 7 depicts the AR frequencies during EPRs (Fig. 7a) versus climatology (Fig. 7b). The southeastern and mid-Atlantic United States, outside the Appalachians, is affected by an AR approximately 20%–30% of EPR’s lifetimes, far above the climatological percentage of 8%–12%. In addition, every EPR has at least one AR in eastern North America during part of the EPR’s lifetime. The region of most frequent ARs is just to the southeast of the region of greatest precipitation (Fig. 3a) and greatest anomalous southwesterly IVT (Fig. 6a), in a southwest to northeast axis from the northern Gulf of Mexico to the mid-Atlantic United States. This region, along with the mean southwesterly IVT during EPRs (Fig. 6a), indicates that ARs transport abundant moisture from the Gulf of Mexico and/or western Atlantic northward into eastern North America, ascend at their northern and/or western tips in regions of forced lifting, such as from cyclogenesis and frontogenesis, and deposit their moisture as precipitation. This mechanism supports our moisture budget results and the findings of Sodemann and Stohl (2013) for a monthlong wet period in Norway and Mahoney et al. (2016) for shorter-duration heavy precipitation events in the southeastern United States. Like for other synoptic-scale features, the region of highest frequency of ARs shifts eastward throughout the EPR’s lifetime, from the far southwestern part of the eastern North American region at ts to just offshore the East Coast by te (not shown). Many EPRs are characterized by a series of two or three ARs, likely associated with a series of cyclones, occasionally with two separate ARs at the same time (not shown).
(a) EPR frequency of an AR, or the fraction of EPR time steps featuring an AR. The magenta dot marks the composite EPR centroid. (b) Climatological frequency of an AR. The black boundary encloses the eastern North American region.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
We repeat the analysis using two other AR identification algorithms (not shown). Although they lead to different absolute values of AR frequencies (and AR contributions to the moisture fluxes shown in the next subsection), using the other algorithms yields the same overall conclusions, since both algorithms lead to the same relative comparisons to climatology and evolution throughout the EPR’s life cycle.
e. Moisture budget
Figure 8a depicts the components of the moisture budget before and during EPRs. The box and whisker plots show that although there is substantial variability between EPRs, there are still clear patterns in the evolution. The net moisture flux [M in Eq. (14)] and net moisture storage {[S] in Eq. (15)} are below climatology at ts−3 but increase leading up to ts and peak well above climatology at ts. M decreases by tm but is still strongly positive and above climatology before decreasing sharply to near zero by te. The [S] decreases to a slightly negative value by tm and decreases further by te. Precipitation {[P] in Eq. (13)} lags behind, being below climatology before ts but then increasing to above climatology by ts and peaking at around tm before decreasing to climatology by te, consistent with our EPR definition. These changes indicate that moisture is being converged into the region and stored leading up to the EPR (M − [P] > 0), and then precipitation exceeds the moisture convergence slightly during the EPR after ts (M − [P] < 0) and is much greater than the moisture convergence, depleting the moisture, at te (M − [P] ≪ 0). The evaporation term {[E] in Eq. (13)} is small though non-negligible, indicating that most of the moisture used in precipitation generation comes from outside eastern North America, which is not surprising since evaporation over a land region during winter is small. It is nearly constant and near climatology during EPRs, indicating that it is likely not an important contributor to EPRs. Averaged over the EPR’s lifetimes, [P] and M are both above climatology, but [S] is negative, below the climatology of zero, indicating that precipitation depletes the moisture faster than it is replenished through moisture influx (M − [P] < 0).
(a) Terms of the moisture budget (mm day−1) as in Eqs. (13)–(15) as depicted in the legend at various times indicated on the horizontal axis. (b) As in (a), but only for MGofM, M, and P, with cyan, black, and green box-and-whisker plots added, respectively. (c) As in (a), but only for AR contributions to the moisture fluxes. The cyan and black box-and-whisker plots are for MGofM and M, respectively. Circles on the left and right side represent the climatological and EPR mean values, respectively, corresponding to the lines of the same color. The ts and te times are depicted by the left and right dash–dotted lines, respectively. In the box-and-whisker plots, the box boundaries indicate the 25th and 75th percentiles, and the whisker tips indicate the 10th and 90th percentiles. The boxes that appear as straight lines indicate that the 25th and 75th percentiles are identical, and boxes with the 10th percentile line missing indicate that the 10th percentile is identical to the 25th percentile, indicated by the lower bound of the box. In most of these instances, one or more of the percentile values are zero since some EPRs do not have an AR at that particular time, especially before and after the EPR.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
We gain additional insight by partitioning M into the fluxes through the four boundary segments individually and analyzing the AR contributions to the fluxes. Averaged over the EPRs’ lifetimes, there is a near climatological western boundary segment flux [Mwest in Eq. (14)], a much larger than climatological MGofM, a negative and near climatological East Coast flux [MEC in Eq. (14)], and a negative and larger than climatological northern boundary segment flux [Mnorth in Eq. (14)] (Fig. 8a). The AR contributions follow a similar pattern, with those of Mwest and MGofM both being much larger than climatology and both EPR and climatological AR contributions to MEC being small (Fig. 8c). These patterns largely result from the mean southwesterly IVT during EPRs, the meridional water vapor gradient, and the propensity for ARs to cross the Gulf of Mexico boundary segment. In addition, the EPR mean MGofM and its AR contribution are more anomalous and have a sharper increase leading up to ts and decrease at te than the fluxes through the other boundary segments. This evolution is supported by the southerly flow from the Gulf of Mexico at ts and tm transitioning to a more westerly or even northwesterly flow by te depicted in Figs. 5a,c,e. Also, all but two EPRs had an average positive AR contribution to MGofM. Even Mwest could have a partial contribution from air parcels traversing the far western part of the Gulf of Mexico and moving northeastward through the western boundary segment. These results indicate, unsurprisingly, that ARs transporting moisture from the Gulf of Mexico and Caribbean Sea, which are the sources of MGofM, are key in promoting heavy precipitation during EPRs. The small AR contribution to MEC could result from varying AR orientations sometimes causing positive or negative AR contributions to MEC. The box-and-whisker plots in Figs. 8b,c show that while there is substantial variability in M and MGofM, their AR contributions, and precipitation among the EPRs, there is a strong EPR evolution signal in the median and 75th and higher percentiles.
5. Differences between high and low AC EPRs
a. 500-hPa height
Figure 9 depicts a side-by-side comparison of the composite 500-hPa height anomaly and standard deviation anomaly of low AC EPRs (left) and high AC EPRs (right) at tm. It shows that the 500-hPa features in the high AC composite are similar to but stronger than in the low AC composite, though the ridge in eastern North America does not extend as far north in the high AC composite. It also shows that in most of North America, the 500-hPa variability among high AC EPRs is smaller than among low AC EPRs, though it is smaller than climatology among both high and low AC EPRs. We obtain similar comparisons at ts and te (not shown).
(a) Composite low anomaly correlation (AC) EPR 500-hPa height (solid contours; 6-dam interval) and anomaly (shaded according to color bar; dam) at tm. (b) As in (a), but for high AC EPRs. (c) Composite low AC EPR 500-hPa σz,ano (shaded according to color bar; dam). (d) As in (c), but for high AC EPRs. The yellow boundary encloses the eastern North American region. In (a) and (c), the navy blue dot marks the composite low AC EPR centroid. In (b) and (d), the navy blue dot marks the composite high AC EPR centroid.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
b. Thermodynamics and precipitation
Figure 10 depicts a side-by-side comparison of the composite θe,max anomaly, wind anomalies at the level of θe,max, SLP, and precipitation for low AC EPRs (left) and high AC EPRs (right). Similar to the 500-hPa features, the high θe,max anomalies and wind anomalies at the level of θe,max are substantially more pronounced in the high AC composite than in the low AC composite. On average, high AC EPRs have the heaviest precipitation over a region extending from the Deep South of the United States northeastward into eastern Canada, while low AC EPRs have the heaviest precipitation closer to the U.S. East Coast. Averaging over eastern North America, high AC EPRs are slightly wetter than low AC EPRs.
(a) As in Fig. 3a, but for low AC EPRs. (b) As in (a), but for high AC EPRs. (c) As in Fig. 5c, but for the composite EPR time-mean during low AC EPRs. (d) As in (c), but for high AC EPRs. In (a) and (c), the purple and green dots mark the composite low AC EPR centroid. In (b) and (d), the purple and green dots mark the composite high AC EPR centroid.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
c. Surface low and high pressure center frequencies
Figure 11 depicts a comparison of composite cyclone and anticyclone frequency anomalies during low and high AC EPRs. The high cyclone frequency regions during high AC EPRs are farther west, with more signal of a Great Lakes cyclone track and less signal of an East Coast cyclone track, than during low AC EPRs. This is consistent with the precipitation occurring farther inland during high AC EPRs. The high anticyclone frequency regions during low and high AC EPRs are similar, but the subtropical Atlantic anticyclone is more frequent during high AC EPRs. The southwesterly IVT anomalies are also stronger during high AC EPRs than during low AC EPRs.
(a) As in Fig. 6a, but for low AC EPRs. (b) As in (a), but for high AC EPRs. (c) As in Fig. 6b, but for low AC EPRs. (d) As in (c), but for high AC EPRs. In (a) and (c), the magenta dot marks the composite low AC EPR centroid. In (b) and (d), the magenta dot marks the composite high AC EPR centroid.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
d. Moisture budget and ARs
Figure 12 depicts a comparison of the composite AR contributions to the moisture fluxes and AR frequencies between low and high AC EPRs. The AR contribution to MGofM is larger in the high AC composite (Fig. 12b) than in the low AC composite (Fig. 12a). ARs are more frequent and extend farther northwest during high AC EPRs (Fig. 12d) than during low AC EPRs (Fig. 12c). Both of these differences are consistent with the more frequent and stronger subtropical Atlantic anticyclones during high AC EPRs (shown in Fig. 10d and Fig. 11d). The AR contribution to MEC is more negative in the high AC composite than in the low AC composite, possibly due to the lower frequency of East Coast cyclones (shown in Fig. 11b) that transport moisture northward into the northeastern United States from the western Atlantic. This difference almost cancels the larger AR contribution to MGofM in the high AC composite, leading to a similar AR contribution to M in the high AC EPR composite compared to the low AC EPR composite. Like for the composite over all EPRs, the box-and-whisker plots in Figs. 12a,b show that while there is substantial variability in AR contributions to M and MGofM among the EPRs, there is a strong EPR evolution signal in the median and 75th and higher percentiles. The full moisture fluxes show similar patterns (not shown).
(a) As in Fig. 8c, but for low AC EPRs. (b) As in (a), but for high AC EPRs. (c) As in Fig. 7a, but for low AC EPRs. (d) As in (c), but for high AC EPRs. In (c), the magenta dot marks the composite low AC EPR centroid. In (d), the magenta dot marks the composite high AC EPR centroid.
Citation: Monthly Weather Review 150, 7; 10.1175/MWR-D-21-0255.1
6. Conclusions
We define EPRs as persistent periods of extreme precipitation volume, normalized by climatology, and analyze them during the eastern North American winter. We find 62 EPRs, with the number of EPRs decreasing rapidly with increasing duration and the vast majority of EPRs lasting between 5 and 8 days. EPRs impact most of eastern North America with heavy precipitation, with a broad southwest–northeast-oriented axis from Louisiana (LA) to Quebec being the most affected, often with large societal impacts such as flooding. EPR duration appears to be more correlated to persistence rather than intensity of the underlying dynamics and thermodynamics, though there is uncertainty in this conclusion due to the relatively large variability between EPRs.
A composite dynamic and thermodynamic analysis shows that EPRs are generally associated with a persistent anomalous 500-hPa trough in western North America, a persistent anomalous 500-hPa ridge in the central Pacific, and a developing anomalous 500-hPa ridge in eastern North America before the EPR start. The pattern slowly moves eastward as the EPR progresses, except for the anomalous ridge in the central Pacific, which remains stationary. Along with the favorable 500-hPa pattern, EPRs are also characterized by intrusions of moist, tropical air and a strong baroclinic zone from the central United States northeastward into Atlantic Canada. This pattern is supported by the evolution of SLP and θe,max anomalies. Although there is substantial variability between EPRs, the variability in the 500-hPa pattern over North America among EPRs, especially high AC EPRs, is still lower than climatology.
EPRs are also associated with southwest–northeast-oriented axes of anomalously high cyclone frequency in the midwestern United States and along the U.S. East Coast, corresponding to cyclone tracks favorable for heavy precipitation, especially the Colorado low track, consistent with Mercer and Richman (2007) and Bentley et al. (2019). Regions of anomalously high anticyclone frequency to the north and south increase the baroclinicity and moisture transport from the south.
During EPRs, there are anomalously higher frequencies of ARs in all categories, particularly in the southern and eastern United States, supporting the results of Sodemann and Stohl (2013) and Mahoney et al. (2016). The ARs during EPRs also tend to be located northwest of their climatological locations and feed into the cyclones at their northern ends, where they deposit their moisture as precipitation. There is a large net influx of moisture into eastern North America leading to a net positive moisture storage before and at the EPR start. Then, heavy precipitation balances out or slightly exceeds the moisture influx until the EPR end, when a net outward moisture flux and remaining precipitation leads to a net negative moisture storage, and the atmosphere dries out, ending the precipitation. The Gulf of Mexico flux is large, positive, and is the most anomalous of the fluxes through the four boundary segments, with a substantial portion of it being associated with ARs, indicating the importance of ARs transporting moisture from the Gulf of Mexico and Caribbean Sea during EPRs.
High AC EPRs have similar 500-hPa features to low AC EPRs over North America, but they are generally stronger, though with the ridge in eastern North America not extending as far north. The synoptic-scale weather patterns among high AC EPRs show less variability than among low AC EPRs. The warmth and moisture anomalies, moisture flux (particularly from the Gulf of Mexico), AR frequencies, and inland precipitation over eastern North America are substantially greater in the high AC composite than in the low AC composite. High AC EPRs are also associated with a farther inland cyclone track, higher frequency of subtropical Atlantic anticyclones, and slightly lower coastal precipitation than low AC EPRs.
The EPR composite synoptic-scale 500-hPa and SLP patterns are roughly similar to those found in the northwestern United States (Lackmann and Gyakum 1996) and northwestern Canada (Lackmann and Gyakum 1999) for short-duration events. However, unlike Lackmann and Gyakum (1996) and Lackmann and Gyakum (1999), the EPR composite has the negative 500-hPa and SLP anomalies to the west being weaker than the positive anomalies to the east, which could indicate the importance of the subtropical western Atlantic anticyclone during EPRs. Lackmann and Gyakum (1996) also found these patterns farther south relative to the region being studied, perhaps due to the high latitude of northwestern Canada and the effects of the Rockies, but also found substantial case-to-case variability, similar to our EPRs. Similarly, Moore et al. (2020) also found that long-duration extreme precipitation events in northern California were associated with series of cyclones, and Jennrich et al. (2020); Sodemann and Stohl (2013) found that similar events were associated with transports of moist, subtropical or tropical air into the regions being studied, often associated with ARs feeding into cyclones, generating heavy precipitation.
Our study has a few limitations. First, owing to the EPR definition, long periods of extreme precipitation with a break of 2 or more days are not considered, though such periods can also lead to severe flooding. Second, like with many reanalyses, the ERA5 precipitation may not be accurate with convective precipitation, but is likely satisfactory for our study as discussed earlier. Finally, perhaps the biggest limitation of this study is the substantial variability between EPRs, despite the large-scale composite anomalies being robust. However, as discussed earlier, the EPR synoptic-scale 500-hPa variability is still lower than climatology, especially among high AC EPRs.
Acknowledgments.
We thank the three anonymous reviewers, Ron McTaggart-Cowan, and Shawn Milrad for their helpful feedback, which has substantially improved the manuscript. To detect ARs, we used the source code from Pan and Lu (2019) with a few slight modifications. This work has been supported by the Fonds de recherche du Québec-Nature et technologies (FRQNT) through the first author’s Master’s and Doctoral Research Scholarship, an NOAA NGGPS (NA16NWS4680018) Grant, and the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant.
Data availability statement.
The ERA5 data used in this study are provided by the Copernicus Climate Change Service at https://cds.climate.copernicus.eu/ (Hersbach et al. 2020). The land–sea mask used to exclude ocean areas is provided by the National Center for Atmospheric Research at http://www.ncl.ucar.edu/Applications/Data/cdf/landsea.nc.
REFERENCES
Archambault, H. M., L. F. Bosart, D. Keyser, and A. R. Aiyyer, 2008: Influence of large-scale flow regimes on cool-season precipitation in the northeastern United States. Mon. Wea. Rev., 136, 2945–2963, https://doi.org/10.1175/2007MWR2308.1.
Archambault, H. M., D. Keyser, and L. F. Bosart, 2010: Relationships between large-scale regime transitions and major cool-season precipitation events in the northeastern United States. Mon. Wea. Rev., 138, 3454–3473, https://doi.org/10.1175/2010MWR3362.1.
Beck, H. E., and Coauthors, 2019: Daily evaluation of 26 precipiation datasets using stage-IV gauge-radar data for the CONUS. Hydrol. Earth Syst. Sci., 23, 207–224, https://doi.org/10.5194/hess-23-207-2019.
Bentley, A. M., L. F. Bosart, and D. Keyser, 2019: A climatology of extratropical cyclones leading to extreme weather events over central and eastern North America. Mon. Wea. Rev., 147, 1471–1490, https://doi.org/10.1175/MWR-D-18-0453.1.
Bosart, L. F., 1981: The Presidents’ Day snowstorm of 18–19 February 1979: A subsynoptic-scale event. Mon. Wea. Rev., 109, 1542–1566, https://doi.org/10.1175/1520-0493(1981)109<1542:TPDSOF>2.0.CO;2.
Changnon, S. A., J. P. Lardner, J. L. Vogel, and P. G. Vinzani, 1983: Floods of winter-spring 1982 in Illinois. Report of Investigation 105, Illinois State Water Survey, 65 pp., https://www.ideals.illinois.edu/bitstream/handle/2142/77809/ISWSRI-105.pdf.
Changnon, S. A., K. E. Kunkel, and D. Changnon, 2008: Winter 2007–2008: Record-setting storms caused major damages in Illinois. Tech. Rep., Illinois State Water Survey, 50 pp., http://www.isws.illinois.edu/pubdoc/DCS/ISWSDCS2008-02.pdf.
Crossett, C. C., A. K. Betts, L.-A. L. Dupigny-Giroux, and A. Bomblies, 2020: Evaluation of daily precipitation from the ERA5 global reanalysis against GHCN observations in the northeastern United States. Climate, 8, 148, https://doi.org/10.3390/cli8120148.
Doswell, C. A., C. Ramis, R. Romero, and S. Alonso, 1998: A diagnostic study of three heavy precipitation episodes in the Western Mediterranean Region. Wea. Forecasting, 13, 102–124, https://doi.org/10.1175/1520-0434(1998)013<0102:ADSOTH>2.0.CO;2.
Eckhardt, S., A. Stohl, H. Wernli, P. James, C. Forster, and N. Spichtinger, 2004: A15-year climatology of warm conveyor belts. J. Climate, 17, 218–237, https://doi.org/10.1175/1520-0442(2004)017<0218:AYCOWC>2.0.CO;2.
Fish, M. A., A. M. Wilson, and F. M. Ralph, 2019: Atmospheric river families: Definition and associated synoptic conditions. J. Hydrometeor., 20, 2091–2108, https://doi.org/10.1175/JHM-D-18-0217.1.
Gyakum, J. R., 2008: The application of Fred Sanders’ teaching to current research on extreme cold-season precipitation events in the Saint Lawrence River Valley Region. Synoptic-Dynamic Meteorology and Weather Analysis and Forecasting: A Tribute to Fred Sanders, Meteor. Monogr., No. 55, 241–250, https://doi.org/10.1175/0065-9401-33.55.241.
Gyakum, J. R., and P. J. Roebber, 2001: The 1998 ice storm—Analysis of a planetary-scale event. Mon. Wea. Rev., 129, 2983–2997, https://doi.org/10.1175/1520-0493(2001)129<2983:TISAOA>2.0.CO;2.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Jennrich, G. C., J. C. Furtado, J. B. Basara, and E. R. Martin, 2020: Synoptic characteristics of 14-day extreme precipitation events across the United States. J. Climate, 33, 6423–6440, https://doi.org/10.1175/JCLI-D-19-0563.1.
Knippertz, P., and H. Wernli, 2010: A Lagrangian climatology of tropical moisture exports to the Northern Hemisphere extratropics. J. Climate, 23, 987–1003, https://doi.org/10.1175/2009JCLI3333.1.
Lackmann, G. M., and J. R. Gyakum, 1996: The synoptic- and planetary-scale signatures of precipitating systems over the Mackenzie River Basin. Atmos.–Ocean, 34, 647–674, https://doi.org/10.1080/07055900.1996.9649581.
Lackmann, G. M., and J. R. Gyakum, 1999: Heavy cold-season precipitation in the northwestern United States: Synoptic climatology and an analysis of the flood of 17–18 January 1986. Wea. Forecasting, 14, 687–700, https://doi.org/10.1175/1520-0434(1999)014<0687:HCSPIT>2.0.CO;2.
Lackmann, G. M., J. R. Gyakum, and R. Benoit, 1998: Moisture transport diagnosis of a wintertime precipitation event in the Mackenzie River basin. Mon. Wea. Rev., 126, 668–692, https://doi.org/10.1175/1520-0493(1998)126<0668:MTDOAW>2.0.CO;2.
Lavers, D. A., and G. Villarini, 2013: Atmospheric rivers and flooding over the central United States. J. Climate, 26, 7829–7836, https://doi.org/10.1175/JCLI-D-13-00212.1.
Lavers, D. A., and G. Villarini, 2015: The contribution of atmospheric rivers to precipitation in Europe and the United States. J. Hydrol., 522, 382–390, https://doi.org/10.1016/j.jhydrol.2014.12.010.
Lenggenhager, S., M. Croci-Maspoli, S. Brönnimann, and O. Martius, 2019: On the dynamical coupling between atmospheric blocks and heavy precipitation events: A discussion of the southern Alpine flood in October 2000. Quart. J. Roy. Meteor. Soc., 145, 530–545, https://doi.org/10.1002/qj.3449.
Mahoney, K., and Coauthors, 2016: Understanding the role of atmospheric rivers in heavy precipitation in the southeast United States. Mon. Wea. Rev., 144, 1617–1632, https://doi.org/10.1175/MWR-D-15-0279.1.
Mailier, P. J., D. B. Stephenson, C. A. T. Ferro, and K. I. Hodges, 2006: Serial clustering of extratropical cyclones. Mon. Wea. Rev., 134, 2224–2240, https://doi.org/10.1175/MWR3160.1.
Mercer, A. E., and M. B. Richman, 2007: Statistical differences of quasigeostrophic variables, stability, and moisture profiles in North American storm tracks. Mon. Wea. Rev., 135, 2312–2338, https://doi.org/10.1175/MWR3395.1.
Miller, J., 1946: Cyclogenesis in the Atlantic coastal region of the United States. J. Meteor., 3, 31–44, https://doi.org/10.1175/1520-0469(1946)003<0031:CITACR>2.0.CO;2.
Milrad, S. M., E. H. Atallah, and J. R. Gyakum, 2009: Synoptic-scale characteristics and pre-cursors of cool-season precipitation events at St. John’s, Newfoundland, 1979–2005. Wea. Forecasting, 24, 667–689, https://doi.org/10.1175/2008WAF2222167.1.
Milrad, S. M., E. H. Atallah, and J. R. Gyakum, 2010a: A diagnostic examination of consecutive extreme cool-season precipitation events at St. John’s, Newfoundland, in December 2008. Wea. Forecasting, 25, 997–1026, https://doi.org/10.1175/2010WAF2222371.1.
Milrad, S. M., E. H. Atallah, and J. R. Gyakum, 2010b: Synoptic typing of extreme cool-season precipitation events at St. John’s, Newfoundland, 1979–2005. Wea. Forecasting, 25, 562–586, https://doi.org/10.1175/2009WAF2222301.1.
Milrad, S. M., E. H. Atallah, and J. R. Gyakum, 2013: Precipitation modulation by the Saint Lawrence River Valley in association with transitioning tropical cyclones. Wea. Forecasting, 28, 331–352, https://doi.org/10.1175/WAF-D-12-00071.1.
Milrad, S. M., J. R. Gyakum, and E. H. Atallah, 2015: A meteorological analysis of the 2013 Alberta flood: Antecedent large-scale flow pattern and synoptic–dynamic characteristics. Mon. Wea. Rev., 143, 2817–2841, https://doi.org/10.1175/MWR-D-14-00236.1.
Moore, B. J., K. M. Mahoney, E. M. Sukovich, R. Cifelli, and T. M. Hamill, 2015: Climatology and environmental characteristics of extreme precipitation events in the southeastern United States. Mon. Wea. Rev., 143, 718–741, https://doi.org/10.1175/MWR-D-14-00065.1.
Moore, B. J., D. Keyser, and L. F. Bosart, 2019: Linkages between extreme precipitation events in the central and eastern United States and Rossby wave breaking. Mon. Wea. Rev., 147, 3327–3349, https://doi.org/10.1175/MWR-D-19-0047.1.
Moore, B. J., A. B. White, D. J. Gottas, and P. J. Neiman, 2020: Extreme precipitation events in Northern California during winter 2016–17: Multiscale analysis and climatological perspective. Mon. Wea. Rev., 148, 1049–1074, https://doi.org/10.1175/MWR-D-19-0242.1.
Moore, B. J., A. B. White, and D. J. Gottas, 2021: Characteristics of long-duration heavy precipitation events along the West Coast of the United States. Mon. Wea. Rev., 149, 2255–2277, https://doi.org/10.1175/MWR-D-20-0336.1.
NOAA/NCEI, 2016: State of the climate: National Climate Report for December 2015. NOAA/National Centers for Environmental Information, accessed 1 February 2022, https://www.ncdc.noaa.gov/sotc/national/201512.
NOAA/NCEI, 2018: State of the climate: National Climate Report for February 2018. NOAA/National Centers for Environmental Information, accessed 1 February 2022, https://www.ncdc.noaa.gov/sotc/national/201802.
NOAA/NCEI, 2019: State of the climate: National climate report for February 2019. NOAA/National Centers for Environmental Information, accessed 1 February 2022, https://www.ncdc.noaa.gov/sotc/national/201902.
Novak, D. R., C. Bailey, K. F. Brill, P. Burke, W. A. Hogsett, R. Rausch, and M. Schichtel, 2014: Precipitation and temperature forecast performance at the Weather Prediction Center. Wea. Forecasting, 29, 489–504, https://doi.org/10.1175/WAF-D-13-00066.1.
Pan, M. P., and M. Lu, 2019: A novel atmospheric river identification algorithm. Water Resour. Res., 28, 6039–6053, https://doi.org/10.1029/2018WR024407.
Peterson, T. C., and M. O. Baringer, 2009: State of the climate in 2008. Bull. Amer. Meteor. Soc., 90 (8), S1–S196, https://doi.org/10.1175/BAMS-90-8-StateoftheClimate.
Pfahl, S., and H. Wernli, 2012: Quantifying the relevance of cyclones for precipitation extremes. J. Climate, 25, 6770–6780, https://doi.org/10.1175/JCLI-D-11-00705.1.
Pfahl, S., E. Madonna, M. Boettcher, H. Joos, and H. Wernli, 2014: Warm conveyor belts in the ERA-Interim dataset (1979–2010). Part II: Moisture origin and relevance for precipitation. J. Climate, 27, 27–40, https://doi.org/10.1175/JCLI-D-13-00223.1.
Pinto, J. G., I. Gómara, G. Masato, H. F. Dacre, T. Woollings, and R. Caballero, 2014: Large-scale dynamics associated with clustering of extratropical cyclones affecting Western Europe. J. Geophys. Res. Atmos., 119, 13 704–13 719, https://doi.org/10.1002/2014JD022305.
Priestley, M. D. K., J. G. Pinto, H. F. Dacre, and L. C. Shaffrey, 2017a: The role of cyclone clustering during the stormy winter of 2013/2014. Weather, 72, 187–192, https://doi.org/10.1002/wea.3025.
Priestley, M. D. K., J. G. Pinto, H. F. Dacre, and L. C. Shaffrey, 2017b: Rossby wave breaking, the upper level jet and serial clustering of extratropical cyclones in Western Europe. Geophys. Res. Lett., 44, 514–521, https://doi.org/10.1002/2016GL071277.
Ralph, F. M., J. J. Rutz, J. M. Cordeira, and M. Dettinger, 2019: A scale to characterize the strength and impacts of atmospheric rivers. Bull. Amer. Meteor. Soc., 100, 269–289, https://doi.org/10.1175/BAMS-D-18-0023.1.
Sauer, V. B., and J. M. Fulford, 1983: Floods of December 1982 and January 1983 in central and southern Mississippi River basin. Tech. Rep., U.S. Geological Survey, 46 pp., https://doi.org/10.3133/ofr83213.
Sharma, S., and Coauthors, 2017: Eastern U.S. verification of ensemble precipitation forecasts. Wea. Forecasting, 32, 117–139, https://doi.org/10.1175/WAF-D-16-0094.1.
Sodemann, H., and A. Stohl, 2013: Moisture origin and meridional transport in atmospheric rivers and their association with multiple cyclones. Mon. Wea. Rev., 141, 2850–2868, https://doi.org/10.1175/MWR-D-12-00256.1.
Starr, V. P., and J. P. Peixoto, 1958: On the global balance of water vapor and the hydrology of deserts. Tellus, 10, 189–194, https://doi.org/10.3402/tellusa.v10i2.9237.
Tarek, M., F. P. Brissette, and R. Arsenault, 2020: Evaluation of the ERA5 reanalysis as a potential reference dataset for hydrological modelling over North America. Hydrol. Earth Syst. Sci., 24, 2527–2544, https://doi.org/10.5194/hess-24-2527-2020.
Wagner, A. J., 1982: The weather and circulation of January 1982—A stormy month with two record cold waves. Mon. Wea. Rev., 110, 310–317, https://doi.org/10.1175/1520-0493(1982)110<0310:WACOJ>2.0.CO;2.
Wallace, J. M., G.-H. Lim, and M. L. Blackmon, 1988: Relationship between cyclone tracks, anticyclone tracks and baroclinic waveguides. J. Atmos. Sci., 45, 439–462, https://doi.org/10.1175/1520-0469(1988)045<0439:RBCTAT>2.0.CO;2.
Wang, X. L., V. R. Swail, and F. W. Zwiers, 2006: Climatology and changes of extratropical cyclone activity: Comparison of ERA-40 with NCEP–NCAR reanalysis for 1958–2001. J. Climate, 19, 3145–3166, https://doi.org/10.1175/JCLI3781.1.
Wilks, D. S., 2016: “The stippling shows statistically significant grid points”: How research results are routinely overstated and overinterpreted, and what to do about it. Bull. Amer. Meteor. Soc., 97, 2263–2273, https://doi.org/10.1175/BAMS-D-15-00267.1.