1

Topic 4.2

ADVANCES IN UNDERSTANDING ET

Rapporteur: John Gyakum Professor and Chair Department of Atmospheric and Oceanic Sciences McGill University Canada Email: john.gyakum@mcgill.ca Phone: +1-514-398-3760 Rapporteur: Julia Keller Deutscher Wetterdienst Research and Development – Numerical Models Frankfurter Str. 135 63067 Offenbach Germany Email: julia.keller@dwd.de Phone: +49-170-5556-448 Working Group Members: Bob Hart, Chris Davis, Chris Fogarty, Christian Grams, Clark Evans, Elizabeth Ritchie, Fermin Elizaga, Florian Pantillon, Heather Archambault, Jenni Evans, João Rafael Dias Pinto, Kimberly Wood, Kyle Griffin, Lance Bosart, Matthew Kucas, Naoko Kitabatake, Pat Harr, Ron McTaggart-Cowan, Shawn Milrad, Sim Aberson (Names in red correspond to the contributors) Abstract: This report summarizes the current state of research on the extratropical transition (ET) of tropical cyclones (TC). It describes new climatological aspects of ET in several ocean basins, advances in forecasting ET and the progress that has been made in getting a more complete understanding of the processes involved during the ET and the interaction of the TC with the midlatitude flow. 4.2.1 Introduction The processes associated with the transformation of a tropical cyclone (TC), as it travels poleward onto the extratropical westerlies, are typically referred to as the Extratropical Transition (ET). The interaction of the transitioning TC and the extratropical westerlies may have strong impacts on both the ET system itself as well as on the midlatitude flow configuration. The need to study the ET system itself is particularly compelling, for several reasons. First, the ET’s areal coverage of wind and rain typically increases from its TC phase. Second, the ET, as it interacts with the baroclinic westerlies, may re-intensify and produce TC-strength winds over increasingly large areas during its poleward movement. Third, the combination of the ET’s tropical air mass and the midlatitude triggers for ascent, often produce copious rains and life-threatening flooding. Finally, the ET often travels into populated coastal regions with significant vulnerability. The detailed processes that determine the impact of the ET system on the extratropical westerlies are another important aspect that requires further investigation. Such a modification of the extratropical flow on the one hand may trigger the development of high impact weather events in regions lying far downstream of the ET itself. On the other hand they often coincide with reduced predictability in exactly those downstream regions that might be affected by the high impact weather events.

2

This report comprises a summary of contributions from WG members who are all working towards an enhanced understanding of the ET. The first part of the report deals with the definition of ET and the progress in developing methods to classify ET, as a standardized objective classification is still lacking. The second part discusses the climatological perspective on ET in several ocean basins and on its downstream impact. Recent operational and research aspects on forecasting ET and the communication of forecasts are provided in section three. Advancements in understanding the detailed process during ET, based on numerous case studies, are provided in section four. The report closes by pointing to open questions and stating the conclusions. 4.2.2 Tools for classification

4.2.2.1 Based on environment Differences in surrounding synoptic environments between dissipating and reintensifying ET systems are used to develop a predictive technique for ET intensity change that can be used to enhance the standard numerical guidance (Felker et al., 2011). Using a set of all historical transitioning TCs between 2000 and 2008 in the western North Pacific, common differences between 850-hPa potential temperature fields surrounding ET intensifiers and ET dissipaters, respectively, were identified. These features were then used as inputs into a support vector machine classification system to create a robust prediction system. Once the system was trained on a random subset of the data (80%), performance was tested on the remaining test set (20%). Overall, it was found that the prediction system was able to correctly predict ET intensity outcome in 75% of the test cases at 72 h prior to ET. 4.2.2.2 Based on CPS Pathway In a study of 82 recurving TCs in the western North Pacific and North Atlantic during 2003-2006, Kofron et al. (2010a) described five major pathways: 1) TCs that interacted with a pre-existing upper-level trough to the northwest during ET, underwent a cold-core transition, and intensified or strengthened (33); 2) TCs that interacted with a pre-existing upper-level trough to the northwest during ET, underwent a warm-seclusion transition, and intensified or strengthened (9); 3) TCs that that rapidly accelerated into a pre-existing upper-level trough to the northeast during ET and intensified or strengthened (16); 4) TCs that underwent extratropical transition and then dissipated (8); and 5) TCs that recurved but did not undergo ET (16). The physical evolution following each pathway is quite distinct. The question of how to identify quantitatively and objectively which pathway a recurving TC is following is then addressed. Several quantities, which have been previously proposed for ET, are investigated to determine whether they exhibit consistent behaviour that can be used to correctly identify ET. These parameters include: frontogenesis; 500-hPa geopotential height open wave; and cyclone phase-space parameters in the Navy’s Operational Global Assimilation and Prediction System (NOGAPS). The results show that the 500-hPa geopotential height open wave correctly identifies 81 of 82 cases of recurving TCs, but fails to discriminate between recurving TCs that undergo ET and those that do not undergo ET. The 2-D scalar frontogenesis distinguishes 77 of 82 cases, but also does not discriminate between recurving TCs that undergo ET and those that do not undergo ET. Cyclone phase space distinguishes 81 of 82 cases by the “symmetry parameter”, but is only successful at capturing transitioning ET and recurving non-ET cases properly in 60 out of 82 cases. Thus, none of the quantities were found to be useful classifiers for ET. Isentropic potential vorticity (IPV) is investigated as a quantitative and objective identifier for the different recurving TC pathways in a follow up study (Kofron et al. (2010b). It is found that IPV on the 330-K isentropic surface is a good discriminator of ET (e.g., Figure 1).

3

Figure 1. Averaged 330-K IPV for each pathway centred on the time that the minimum IPV occurs. Note the spread in IPV values at times after "0" that well distinguishes those cases that undergo ET and re-intensify from either "non-ET" or "ET

dissipating" cases. [Adapted from Figure 8b Kofron et al. (2010b)] An “ET time” is identified as the time of minimum 330-K IPV averaged in a 500 km radius of the TC centre. It is found that all 82 recurving TCs meet this “minimum” criterion; however, the evolution of the 300-K IPV is quite distinctive for each pathway following this time (Figure 2) and it is found that ET completion occurs when the IPV exceeds a threshold value of 1.6 PVU with a 94.3% success and 27.6% false alarm rate. (Ritchie)

Figure 2. Composite isentropic potential vorticity every 12 hours for the 48 hours prior to 48 hours post “ET time” as

determined using minimum 330-k IPV: a) cold-core upstream trough ET; b) warm-core seclusion; c) cold-core downstream trough ET; d) ET dissipators; and e) recurving non-ET TCs. All images are scaled from 0.8 – 2.5 PVU.

4

4.2.3 Climatological Aspects

4.2.3.1 Western North Pacific Ocean Climatological characteristics of ETs in the western North Pacific between 1979 and 2004 are examined by using a reanalysis dataset (Japanese Re-Analysis 25 years: JRA-25) at 1.25° × 1.25° horizontal resolution and a criterion of ET defined by indices of the cyclone phase space (CPS; Hart 2003). Although a bogus vortex is implemented for a TC to construct a warm-core structure in the JRA-25 dataset, 48% of all TCs that completed ET were transformed into the cold-core structure before the termination of the bogus implementation (Kitabatake 2010). This result confirms that the use of the JRA-25 dataset for climatological studies on ET is appropriate. Based on this dataset and criteria, 274 (40%) out of 687 TCs completed ET in 26 years. The ratio of TCs that complete ET is about 30% in the months of June-August and nearly 60% in September and October. Although the most frequent latitudes of ET were 35–40°N, several TCs completed ET far north of these latitudes (e.g., north of ~45°N in October, Figure 3). It is considered that they include cases in which the CPS failed to separate a post-ET reintensification stage with the warm-core development from the extratropical transformation stage. Several intense TCs having central pressure below 960 hPa complete ET in autumn. The difference of baroclinic stability in the atmospheric environment and SST are considered to be important factors that cause the seasonal variation of ETs. 4.2.3.2 Eastern North Pacific Ocean Since IWTC-VII, multiple studies have been published that document climatologies of ET in previously under-researched basins, such as the southwestern Indian Ocean (SWIO; Griffin and Bosart 2014) and eastern North Pacific (ENP).

Figure 3. 26-year monthly mean field of Eady growth rate (contour, day-1), SST (colour, °C), and the location of ET completion (circle) in August (left) and October (right) (Kitabatake 2011)

Wood and Ritchie (2012) presented a case study of Tropical Storm Ignacio (1997), an open-ocean ET event in the eastern North Pacific. Though a weak TC during the tropical phase of its lifetime, it interacted with a trough, subsequently reintensified after ET due to favourable baroclinic conditions, and produced measurable precipitation to the northwestern United States. This was followed by a study of ENP ET over the period 1971-2012, in which it was found that ET occurs for only 9% of ENP TCs, as the subtropical ridge in place for much of the season and the strong sea surface temperature gradient prevent most TCs from reaching the baroclinic zone and beginning the ET process (Wood and Ritchie 2014a). Those ENP TCs that do undergo ET often follow an unconventional evolution through cyclone phase space (CPS) compared to the North Atlantic and western North Pacific, retaining their thermal symmetry while losing their warm

5

core prior to completing ET rather than retaining the warm core and becoming more asymmetric (e.g., Figure 4). Such an evolution is not limited to the ENP; Kitabatake (2011) noted this alternate path through CPS for 16.8% of western North Pacific ET cases. Peak ENP ET activity occurs in September and October, offset from peak TC activity in July and August due to increasing frequency of midlatitude troughs and a weaker subtropical ridge later in the TC season. The frequency of ENP ET increases during developing El Niño events, likely due to the warmer sea surface temperatures and more southward-protruding midlatitude troughs. However, ET frequency does not appear to be affected by the Pacific decadal oscillation. The Wood and Ritchie (2014a) study initially used ECMWF 40-year Reanalysis (ERA-40) and ERA-Interim data to perform CPS analyses, but they discovered discrepancies in the depiction of the lower level warm core in the ECMWF fields compared to Final Analysis (FNL) and Japanese 55-year Reanalysis (JRA-55) data (e.g., Figure 5). These differences were also documented in a study that compared TC position and intensity in five reanalyses (Schenkel and Hart 2012). Of the three basins examined, they found that the greatest differences in position and weakest intensities occurred in the ENP, though some of this error was ameliorated in datasets that applied either vortex relocation (CFSR) or wind profile retrievals (JRA-25) to better represent TCs. Further work is being performed in order to quantify the differences in TC representation between seven commonly used global reanalyses (e.g., Wood and Ritchie 2014b). (Wood)

Figure 4. Average CPS frequency per TC during 2001-2010 in the (a) ENP (150 TCs) and (b) ATL (174 TCs). Figure adapted from Wood and Ritchie (2014a)

Figure 5: Zonal geopotential height anomalies from three reanalyses at 0000 UTC 25 Sep 2011 for Hurricane Hilary (115 kt). Figure adapted from Wood and Ritchie (2014a)

6

4.2.3.3 Southwest Indian Ocean While most ocean basins with TC activity have well-documented climatologies of ET published in the last 15 years, the Southwest Indian Ocean (SWIO) has seen few TC climatologies published. The first such climatology was presented by Valadon (1992), which documented the 1960-1989 seasons and found an average of 9 TCs per year, but does not discuss ET. A more recent SWIO TC climatology is presented by Mavume et al. (2009), finding an average of 12.5 TCs per year over the 1980-2007 period. Ramsay et al. (2012) mention the existence of many recurring tracks in the historical records across the southern Indian Ocean, especially over the southwestern potion of the basin, but does not explicitly mention ET either. Griffin and Bosart (2014) presented the first climatology of ET events in the SWIO in the modern satellite era. Figure 6 shows a summary of TC and ET events in the SWIO by (Figure 7a) TC season, and by (Figure 7b) month. On average, 9.4 TCs of tropical storm strength (10-minute sustained winds >34 kt) or higher occur per TC season (total number of TCs is 235). Four TCs undergo ET every year, leading to an average ET rate of 43.8% (Figure 3a).This yields a 44% ET rate, comparable to the 46% rate found by Hart and Evans (2001) for the North Atlantic and the second-highest rate of any TC basin worldwide where ET is commonly observed (27% in the northwest Pacific from Klein et al. (2000); 32% in the south Pacific from Sinclair (2002)). The 9.4 count of annual TC events roughly agrees with the ~10 TC events per year derived from the previous SWIO TC climatology (1960-1989) computed by Valadon (1992). These ET events in the SWIO occur almost exclusively between November and April (Figure 3b) with the highest frequency (~32) in January and February and the lowest frequency (~4) in November and December. The highest ET rate (~57% of all TCs) occurs in January, decreasing to 48%, 48%, and 42% for February, March, and April, respectively (Figure 3b). These late-season SWIO ET rates are similar to the rates calculated by Hart and Evans (2001) for the peak ET months of September and October in the North Atlantic. This suggests the SWIO typically contributes a notable proportion of the annual global ET events. The general lack of landmasses poleward of 25˚S makes the Indian Ocean a prime location for ET, but also means that few ET events have a significant downstream impact on any populated areas with Australia also not extending poleward of 35˚S. (Griffin & Bosart).

Figure 6. Summary of TC and ET events in the SWIO west of 90˚E by (a) TC season and (b) month, 1989-2013. Full height of bar represents TC events, with lower (blue) portion of bar represent the number of TCs undergoing ET. In (a), year on chart

refers to year season ended. In (b), cross-month events sorted by which month contained the end of TC life. Lower (blue) portion of bar represent the number of TCs beginning ET that month.

4.2.3.4 North Atlantic Ocean Hurricane Sandy (2012) and numerous other historical cases (e.g. 1938 Long Island/New England Hurricane) have illustrated that proximity between TC and extratropical cyclones (XCs) can have a substantial impact on the motion of the TC. Here we address the climatology and motion impact of such events over nearly 1.5 centuries. Since an official database of extratropical cyclone tracks has not been established, a gridded dataset was chosen to create this track

a) b)

7

database: the NOAA/CIRES 20th Century Reanalysis (Compo et al. 2011). While this isn’t the highest resolution spatial gridded dataset available, it is the longest in duration and has been utilized for many other similar studies (Truchelut and Hart 2011; Truchelut, Hart, and Luthman 2013). Thus, the first step was to track closed cyclones in this dataset from 1871 through 2012, using the 500mb pressure level. Once this database was created, the locations of tracked cyclones had to be filtered. Given the grid spacing of the dataset, actual TCs might be represented in the dataset, but offset by a small distance from reality. Since TC-TC proximity was not the focus of this part of the study, all best-track TCs within 600km of the 20th Century Reanalysis cyclones were discounted as the same cyclone. This threshold of 600km was chosen because a very, very small percentage of actual TC-TC proximity from Part 1 of the study (not shown) fell within this threshold (i.e., it is extremely rare for two actual TCs to survive within a separation distance of 600km or less). Figure 7 demonstrates graphically the proximity PDF and the thresholds used to filter into the desired subset. Figure 8 illustrates the resulting spatial climatology of TC-XC proximity (in this case 1500km, although other smaller thresholds are also available upon request). This climatology is presented as a function of the TC location (Figure 8a) and the XC location (Figure 8b), and illustrates that TC-XC proximity is most frequent along and just offshore the U.S. East Coast. Thus, while the classic cases of the 1938 New England Hurricane and Sandy (2012) were cases of unusual intensity, the occurrence of the TC-XC proximity was not at all unusual in the context of 142 years of analysis. The mean motion vector for TCs in proximity of diagnosed XCs was calculated, and then compared to the mean TC motion vector regardless of XC proximity. Figure 9 presents this analysis as the anomalous motion from the local mean TC motion. Immediately apparent from Figure 9a is that at 1500km separation distance or less, the aggregate effect of TC-XC interaction is to accelerate TCs near the U.S. East Coast but in a direction parallel to the coast. Thus, the TC motion relative to the coast may not be significantly changed, although the time to landfall may be decreased. However, once the separation distance decreases to 1000km or less (Figure 9b), the anomalous induced motion is clearly onshore, and by 800km separation distance (not shown), the anomalous motion is nearly perpendicular to the shoreline and directed inland. This analysis suggests that at a separation distance of greater than 1000km but less than 1500km, the net effect of a nearby XC is to simply accelerate the TC northeastward by the large-scale trough in which the tracked XC is embedded. In other words, the TC is influenced by the large scale northeast flow of the trough, but not close enough on average to the XC itself to exhibit Fujiwhara interaction with the embedded closed cold-core cyclone. However, once the separation between the TC and the XC decreases to 1000km or less, it is clear that the mean anomalous motion is dominated by the Fujiwhara interaction with the XC and an onshore motion in the northeast U.S. results – since the climatological location of the TC (shown in Figure 8a) and XC (as shown in Figure 8b) position both such that the TC would be driven toward the coast. The mean “axis of rotation” (the centroid between the TC and XC; not shown) is located across the Washington, DC to New York City region, further illustrating the highly populated region under the influence of such interaction, and demonstrated most recently by Hurricane Sandy (2012) but also by earlier classic cases. (Hart)

8

Figure 7. a) Filtering of best track-reanalysis proximity database. b) Example year showing three proximity events

Figure 8. Frequency of occurrence from 1871-2012 of TC-XC proximity within 1500km. a) TC location. b) Extratropical cyclone location (as defined by 500mb closed centre)

Figure 9. Anomalous TC motion (knots) associated with TC-XC proximity. a) within 1500km b) within 1000km

a) b)

a) b)

b) a)

9

4.2.3.5 South Atlantic Ocean Many studies have shown the region of the Atlantic Ocean close the eastern coast of South America (SAO) as a cyclogenetic region (Dias Pinto and da Rocha, 2011). However, over the SAO the occurrence of extratropical transitions ET (or even TT) is still object of debate. Since ET is a process by which a pre-existent warm-core cyclone gradually transforms into a cold-core extratropical cyclone, it is necessary an environment that favours the formation of tropical cyclones in the region. Due to the fact that the average sea surface temperatures (SST) along the Brazilian coast are lower than North Atlantic values and the vertical wind shear is stronger, the genesis of such systems is rare. The lack of research in the SAO not only contributes for a scarce literature on the subject in this part of the basin but it also imposes the study of ET be based on few case studies. Nonetheless, the meteorological community has attracted more attention to the SAO after the first documented hurricane, Catarina, formed close to the Brazilian coast in March 2004 (more details in McTaggart-Cowan et al., 2006). It started as an extratropical cyclone associated with a frontal system and underwent tropical transition (TT), achieving in its later stages full properties of a category 1 hurricane and dissipated over the southern Brazil mainland. Although it did not transition to a cold-core cyclone through the ET process, the genesis of this hurricane led the question on whether these systems formed in the past or will be more common in a changing future climate over the SAO. Cyclone Anita (2010) is another case of unusual event over the SAO. It started as pure subtropical cyclone, evolved to a condition favourable for tropical transition, later developed into a cold-core structure and decayed as an extratropical cyclone (Dias Pinto et al., 2013). During the period favourable for TT, Anita presented a rounded cloudiness structure showing free cloud area in the centre, whose characteristics were also documented during the life cycle of tropical-like storms over the Mediterranean Sea, the Medicanes (Mediterranean+Hurricanes; Emanuel, 2005; Fita et al., 2007). As the Anita approached the Brazilian coast it started interacting with a propagating extratropical cyclone, which increased the baroclinicity of the environment, and lately allowed the extratropical transition to happen. Although Catarina and Anita presented different lifecyles and evolved at the end to different structures, they presented the same precursor environment. Figure 10 shows a simplified view of the life cycle of these two storms by depicting the satellite image of three different instant of their lifecycle. The formation of both Catarina and Anita were associated to a predominately barotropic precursor environment provided by a dipole-blocking-like structure. The occurrence of this structure in mid- and high-levels of troposphere was responsible for the reduction of the vertical wind shear and the westward propagation of these systems. Embedded in such a flow, these systems were able to organize in a structure capable of self-amplification. The occurrence of these two cases is not capable of providing a broader view on the development of tropical cyclones and the possibility of more extratropical transitions. Since this kind of system is not fully documented over the SAO, a complete climatology with improved methods of tracking and classification is required. Besides numerical modelling is also needed to better understand the role of upper-level forcing, provided by the blocking-like flow and the surface fluxes during the transition process. (Pinto)

a) b)

10

Figure 10. Simplified life cycle of two South Atlantic tropical storms. Anita (2010) was the only documented cyclone to undergo ET. Images based on infrared GOES-12 satellite imagery.

4.2.3.6 Downstream Impact a) Precipitation The remnants of eastern North Pacific TCs on the southwestern U.S., mostly in terms of rainfall patterns is investigated in Ritchie et al. (2011). They note that the two patterns “recurving TC” patterns they find tend to bring the majority of the rainfall to southwest U.S. (e.g., Figure 11). Both patterns were associated with an upstream upper-level midlatitude trough that protruded south through a break in the subtropical ridge extended west from central and North America over the Pacific Ocean.

11

Quantitative statistical analysis is used to ascribe the significant TC-related rainfall patterns and large-scale circulation patterns associated with them in the eastern North Pacific in Wood and Ritchie (2013). The database is increased from 42 to 167 TCs over the 21-y period 1989-2009 and EOF and MCA analysis is used to demonstrate the prevalence of a midlatitude trough pattern when TC-related rainfall occurs in the southwestern United States (e.g., Figure 12). Conversely, the presence of a strong subtropical ridge tends to prevent such events from occurring and limits TC-related rainfall to Mexico. Furthermore, the El Niño – Southern Oscillation is also shown to have some effect on TC impacts in the southwestern U.S. as corresponding shifts in the large-scale circulation pattern affect where the TC remnants make landfall. (Ritchie)

Figure 11. Average rainfall swath (mm) for the two recurving TC groups identified in Ritchie et al. (2011) and their

representative members. [Adapted from Figure 5 in Ritchie et al. (2011)].

Figure 12. First MCA pattern of 500-hPa geopotential heights (m) and rainfall pattern (mm) for the positive mode showing interaction of the TC (black dot) with a southward protruding midlatitude trough. Note the similarity of the rainfall pattern

with a combination of groups 1 and 2 in Figure 11.

12

The St. Lawrence River Valley (SLRV) 1,200 km long, (average) 100 km wide valley that stretches through eastern Ontario and Quebec from the east end of Lake Ontario to the Gulf of St. Lawrence. The orography of the SLRV has previously been shown to impact surface winds, observed weather, and precipitation distributions during the cold season. To examine the impact of the SLRV on transitioning tropical cyclones (TCs), we identified 38 cases of named storms whose centres came within 500 km of the SLRV from 1979-2012. We did not a priori consider which stage of extratropical transition (ET) that each TC was in, although no cases were still pure TCs at the time of their closest approach to the SLRV. In 19 of 38 cases (Group A), substantial ageostrophic frontogenesis (> 150 x 10-2 K (100 km)-1 (3 hr)-1 occurs parallel to the long axis of the SLRV at the time of heaviest precipitation (t = 0 h). Five cases (Group B) are null cases (i.e., cases that did not meet the ageostrophic frontogenesis criterion), while 14 cases (Group C) are not analyzed further for various reasons (not shown, detailed in Milrad et al. 2013). A composite of 15 (of 19) Group A cases in which the ageostrophic frontogenesis occurred in the northeastern SLRV (Figure 13a) shows that the ageostrophic frontogenesis is primarily associated with pressure-driven wind channeling (northeasterly winds parallel to the SLRV), with a small contribution from coastal frontogenesis (not shown). In contrast, Group B cases (Figure 13b) show little ageostrophic frontogenesis in the SLRV and are typically associated with either forced channeling or downward momentum transport, in which the near-surface wind direction is similar to the geostrophic wind. The unreliability of NCEP North American Regional Reanalysis (NARR) precipitation in Canada limited our assessment of precipitation distributions to 10 Group A cases from 2003-2011 (when Canadian Precipitation Analysis [CaPA] data were available). In those cases, ageostrophic frontogenesis and enhanced precipitation are concomitant and oriented parallel to the SLRV. A representative Group A case (Frances 2004) is shown in Figures. 13c-d: The heaviest CaPA precipitation and strongest NARR ageostrophic frontogenesis are co-located from the eastern end of Lake Ontario to east of Montreal, with a strong pressure gradient observed along the long axis of the SLRV. A southeast-northwest cross-section for Frances (2004) shows that the ageostrophic frontogenesis is shallow (lowest 75-100 mb of the troposphere) and located primarily within the confines of the SLRV (Figure 13e). Moreover, a NARR sounding (Figure 13f) depicts surface-based cold air, a large near-surface temperature inversion, and moist neutral profiles above the inversion layer. The temperature inversion in Group A cases is primarily due to near-surface cold-air advection (CAA) from northeasterly pressure-driven channeled winds (Figure 10a). There is no preference in Group A cases for synoptic-scale ascent and geostrophic frontogenesis to be oriented purely within and parallel to the long axis of the SLRV. This suggests that the ageostrophic frontogenesis serves as a mesoscale ascent-focusing mechanism for enhanced precipitation in the SLRV.

13

Figure 13. Top: NARR ageostrophic frontogenesis (shaded, x10-2 K (100 km)-1 (3hr)-1) at t = 0 h (time of heaviest precipitation), wherein the ageostrophic wind is the total wind at 30 m minus the geostrophic wind calculated from sea-

level pressure (SLP), and 30 m potential temperature (K, black contours) for composites of (a) 15 ET cases characterized by ageostrophic frontogenesis in the northeastern SLRV and (b) 5 ET cases characterized by no ageostrophic frontogenesis

in the northeastern SLRV. Middle: For Frances (2004), (c) NARR ageostrophic frontogenesis as in panel (a), but total from t = -3 h to t = +3 h, and (d) CaPA total precipitation (mm, shaded) for t = - 6 h to t = + 6 h and NARR SLP (hPa, solid black) at t

= 0 h. The black ovals denote the area of enhanced precipitation in the SLRV. In both (c) and (d), Montreal’s Trudeau International Airport (CYUL) is marked with a blue star and Quebec’s Jean-Lesage International Airport (CYQB) is marked

with a black star. Bottom: Also for Frances (2004), (e) cross-section of ageostrophic frontogenesis (shaded, as in panel (a)) and equivalent potential temperature (K, red contours), at t = 0 h. The crosssectional area is denoted by the black line in

panel (c) and the black star indicates the centre of the SLRV. (f) NARR sounding at t = 0 h at the location of the black star in panel (e). Plotted are temperature (°C, red), dew point (°C, blue), and winds (kts, black barbs).

As such, we propose the following physical pathway to the modulation of heavy precipitation within the SLRV during a Group A event (an associated schematic is shown in Figure 14):

• A transitioning TC approaches the SLRV, and helps to establish a synoptic-scale pressure gradient in the valley, typically pointing from higher pressure northeast to lower pressure southwest (e.g., Figure 13c). The pressure gradient within the SLRV must be > 0.4 hPa (100 km)-1. in order for northeast to southwest pressure-driven wind channeling to start. The channeling results in northeasterly surface winds and CAA along the SLRV (Figure 13a)

14

• The channeled northeasterly surface winds in the SLRV provide a source of low-level cold air, while surface winds outside of the SLRV are primarily southeasterly (e.g., Figure 13a). The combination results in ageostrophic frontogenesis, oriented parallel to the long axis of the SLRV.

• The shallow lower-tropospheric ageostrophic frontogenesis acts as a mesoscale ascent-forcing mechanism that allows air parcels to rise above the inversion to a layer of elevated conditional instability, resulting in larger precipitation totals within the SLRV than outside of it.

In summary, ET cases can mimic cold-season midlatitude cyclones, in that they can establish a pressure gradient within the wide and long SLRV. The subsequent pressure-driven wind channeling leads to ageostrophic frontogenesis, modulating the associated precipitation distribution. This process confines the heaviest precipitation to the populated SLRV corridor within Ontario and Quebec, and presents a large threat to life and property, as ET cases often contain large amounts of tropical moisture that can lead to flash flooding. (Milrad)

Figure 14. Schematic showing the processes involved in producing enhanced precipitation during ET cases in the SLRV. The red circles and dotted lines indicate two common storm locations and tracks. The blue and purple arrows represent the geostrophic and pressure driven channeled surface wind, respectively. The black lines

approximate isobars, with the identified pressure gradient threshold for pressure-driven channeling written in purple. The yellow ellipses are typical regions of ageostrophic frontogenesis and precipitation enhancement within the SLRV.

b) Rossby Wave Trains A statistical climatology of all recurving western North Pacific TCs that undergo ET during 1979–2009 (N = 292) by Archambault et al. (2013) finds that following TC recurvature, significantly amplified flow develops over the North Pacific that persists for ~4 days. The tendency for significantly amplified flow to develop following TC recurvature is sensitive to the strength of the TC–extratropical flow interaction, which is objectively defined by the negative potential vorticity (PV) advection by the irrotational wind associated with the TC. In contrast, it is relatively insensitive to characteristics of the TC, such as its intensity or size, or whether it reintensifies after becoming extratropical.

15

Archambault et al. (2014a) explore the synoptic–dynamic factors underpinning the aforementioned statistical relationships between western North Pacific TC recurvature and extratropical flow amplification. Composite analyses of the recurving western North Pacific TCs identified by Archambault et al. (2013) reveal that a preexisting Rossby wave train migrating from Asia amplifies over the western North Pacific as the TC recurves, and subsequently propagates downstream across the North Pacific and North America into the Gulf of Mexico. In a composite sense, the Rossby wave train traverses ~240 degrees of longitude and lasts ~10 days. Composite analyses of objectively defined strong TC–extratropical flow interactions indicate that strong interactions favour the amplification of a long-lived Rossby wave train that migrates downstream across the North Pacific to North America, and establishes a high-latitude ridge over western North America. In contrast, weak interactions are associated with a short-lived Rossby wave train that decays over the North Pacific without influencing North America.

Figure 15. Composite analyses showing (a) strong (N = 54) and (b) weak (N = 54) TC–extratropical flow interactions at the time of maximum interaction. Analyses show 500-hPa ascent (dashed green, every 2 × 10-3 hPa/s, negative values only),

precipitable water (shaded according to gray scale, mm), and 200-hPa PV (blue, every 1 PVU), irrotational wind (vectors, >2 m/s; purple vectors, >8 m/s), negative PV advection by the irrotational wind (dashed red, every 2 PVU/day starting at −2

PVU/day), and wind speed (shaded according to colour bar, m /s)). The star denotes the point of maximum interaction. The TC symbol denotes the composite TC position. (After Archambault et al. 2014a, their Figure 7).

A comparison of strong and weak TC–extratropical flow interactions at the time of maximum interaction (Figures 15a,b, respectively) reveals that compared to weak interactions, strong interactions feature stronger forcing for ascent and low-level frontogenesis (not shown) that is collated with a region of stronger midlevel ascent over and northeast of the recurving TC in a more moist environment. Strong interactions feature recurving TCs with stronger divergent outflow that impinges upon the eastern flank of a stronger, broader upper-level trough, and are associated with stronger upper-level PV frontogenesis (not shown) within the entrance region of a stronger downstream jet streak. (Archambault) Quinting and Jones (2014) present a climatological investigation of the impact of transitioning TCs on midlatitude Rossby Wave Packets (RWP). Based on the Hilbert transform of the 250 hPa meridional wind (Zimin et al. 2003) they identified RWPs as objects and considered those RWP that occurred seven days before until ten days after ET events in three ocean basins between 1980-2010. A statistically significant increase in RWP amplitude and occurrence frequency associated with ET events over the western North Pacific is found. In the ET relative composite, the RWP occurrence frequency exceeds the climatological mean by up to 20%

16

between 120E and 80W from one day prior to four days after the ET events. The strongest anomaly occurs over the central North Pacific. But even over North America the RWP occurrence frequency exceeds the climatological mean by more than 12%. The statistically significant RWP amplitude increases by up to 2 m /s between 160E and 110 W from ET-time to four days after ET. The observed amplification of the meridional flow is consistent with results of Archambault et al. (2013) who found an increase of the North Pacific flow pattern downstream of ET.

South Indian Ocean TCs that undergo ET have a statistically significant impact on RWPs. However, the deviations from the climatological mean are not as strong as over the western North Pacific. The RWP occurrence frequency exceeds the climatological mean by up to 15% between 60 and 130E from one day prior to ET to two days after ET. The strongest anomaly occurs over the eastern South Indian Ocean about one day after ET. Over the same region, the RWP amplitude is enhanced considerably by up to 2 m /s.

In order to identify synoptic conditions and physical processes that favour the development of RWPs downstream of ET events, the downstream response to western North Pacific ET events is investigated in a composite view. Based on the strength of the Rossby wave signal downstream of each TC, the upper and lower quintile of all ET events are chosen to be representative of cases with and without downstream Rossby wave development. A Hovmöller diagram of the meridional wind suggests that a precursor Rossby wave that emerges from the Asian continent favours the conditions for a Rossby wave train downstream of the ET events. The precursor Rossby wave is strongly amplified during the interaction with the TC and disperses over the North Pacific toward North America. These results coincide with findings of Riemer and Jones (2010). They concluded that the leading edge of a developing baroclinic wave represents an optimal location where an ET system can impact the downstream flow most significantly. Composite maps of Ke budget terms for the RW and noRW cases are created to assess the impact of the TCs on the wave amplification. Ke that is generated through baroclinic conversion in the vicinity of the transitioning TCs, is dispersed via divergence of the ageostrophic geopotential flux. Strong ageostrophic geopotential flux convergence contributes to the development of a primary midlatitude Ke centre on the eastern flank of the trough upstream of the TC in the RW cases. The accumulation of Ke results from the convergence of the midlatitude flow and the TC outflow. This convergence indicates that the TC provides additional Ke to the midlatitude flow and contributes to the Rossby wave amplification. Once a midlatitude Ke centre has developed, Ke is dispersed through div- and converging ageostrophic geopotential fluxes into downstream regions where it contributes to the formation of further Ke centres in the flanks of the developing troughs and ridges. These findings coincide with results from Harr et al. (2000), Harr and Dea (2009) and Keller et al. (2014). In the noRW cases the ageostrophic geopotential flux convergence between the TC outflow and the midlatitude flow, associated with the upstream trough does not occur. Therefore, a Ke centre does not develop in the eastern flank of the upstream trough. Instead, convergence of ageostrophic fluxes due to the merging of the split jet leads to an accumulation of Ke in the crest of the downstream ridge. This accumulation leads to the acceleration of the midlatitude jet but not to an amplification of the RWPs. (Quinting/Keller)

17

4.2.4 Forecasting ET 4.2.4.1 Operational Aspects Since 2010, the Joint Typhoon Warning Center (JTWC) has adapted a few new procedures to diagnosis and predict extratropical transition (ET). On the diagnostic front, JTWC developed a guided technique to streamline cyclone phase analysis (i.e., classify cyclone phase as tropical, subtropical, or extratropical). This process features a “Cyclone Phase Classification Worksheet,” which assesses cyclone phase based on thirteen observable characteristics of a cyclone and its surrounding environment. Although this classification process was originally developed to improve real-time identification and analysis of subtropical cyclones, it may also be applied to analyze extratropical transition. The process is adaptable, so data collected during real-time application will be used to update and improve the worksheet’s classification formula. The worksheet, like many other operational tools and processes in tropical cyclone analysis, is heavily dependent on

Figure 17. As in Figure 16 for anomaly from December to April climatology for the western South Indian Ocean TCs. 114 cases were used.

Figure 16. Anomaly of RWP occurrence frequency [%] (a) and RWP amplitude [m/s] (b) from June to November climatology for western North Pacific TCs. Values that are statistically significant at the 95% confidence level are hatched. Black horizontal bar marks the range of longitudes of TCs at ET-time. White circle marks the mean longitude of all TCs at ET-time. 280 cases were used.

18

scatterometer, imager, and sounder data gathered by microwave sensors carried onboard low-earth orbiting satellites. Further details can be found in Kucas et al. 2014. JTWC forecasters have also integrated two model-based tools to predict ET timing. First, forecasters routinely access the Florida State University (FSU) model-derived cyclone phase space products, which was a relatively new addition to the JTWC extratropical forecasting toolkit four years ago (Hart 2003). Second, forecasters have begun evaluating “storm type” classifications recently added to the Statistical Hurricane Intensity Prediction System (SHIPS) guidance package. The SHIPS algorithm, developed at the Cooperative Institute for Research in the Atmosphere (CIRA), provides cyclone phase space assessments based on storm-centred, deep-layer vertical wind shear and convective instability calculated from GFS-forecasted atmospheric soundings (CIRA-RAMMB). In general, JTWC’s ET forecast process has benefitted from increasing accuracy in global deterministic and ensemble model forecasts over the past four years. Improving simulations of midlatitude flow patterns and dependent tropical cyclone track forecasts have helped forecasters to identify ET transition points and maintain consistent ET timelines from one forecast to the next. Realistic simulations of post-transition structure and intensity in these models have also helped forecasters to formulate more accurate intensity forecasts for the transition period. (Kucas) 4.2.4.2 Data impact studies The value of targeted observations for the forecasts of ET events over the Atlantic is assessed by Anwender et al. (2012). Data denial experiments with the ECMWF Integrated Forecasting System (IFS) showed that the impact of denying data near the transitioning TC is on average as important as is denying data in midlatitude sensitive regions which are determined by extratropical singular vectors. The impact caused by denying the data can be traced to propagate downstream towards Europe, with a maximum degradation at forecast day 4. The degradation results from denying data in a stage before the ET of the TC is completed. In contrast, denying data in midlatitude sensitive regions, strongest forecast degradation occurs at around day 2 and day 4. The loss of information content is larger by denying data in the dynamically active regions in the midlatitudes. Aircraft and satellite radiance data are found the be the most influential data sets in both data denial experiments In a case study for Hurricane Irene (2005), the largest degradations are found for forecasts that were initialised while the TC reached its peak intensity. Denying observations near the storm mostly results in losing the TC in the analysis and subsequent forecast, which enables investigation of the downstream impact of Irene. (Keller) 4.2.4.3 Impact on Predictability For eight recent ET cases the forecast degradation due to ET has been assessed using a novel measure based on the anomaly correlation coefficient (Chapter 6.3 of Grams (2011) and Grams et al. (2014, manuscript in preparation). A first reduction by around 2 days is associated with the ridgebuilding directly downstream of ET. Error in the initial position of the ET system and thus in the track of the ET system result in an even stronger forecast degradation. (Grams) 4.2.4.4 Communication and outreach Advances in understanding ET in recent years have not been limited to traditional scientific research and operational forecasting. Progress is gradually being made in the communication of the impacts from post-tropical cyclones and how those impacts compare/contrast from pure tropical cyclones. For example, in the United States, the National Hurricane Centre (NHC) in recent years have adopted the term “post-tropical” in their bulletins in place of “extratropical” which is a more technical term that is not often understood by the public and media. After 2012’s Hurricane/Post-Tropical Storm Sandy struck the Northeastern United States, a change of practice was also adopted by the NHC to continue issuing tropical-type alerts even for storms that have recently

19

undergone ET. This results in less confusion and encourages citizens to remain vigilant of the destructive nature of some post-tropical storms. The Canadian Hurricane Centre (CHC) has been using the ‘post-tropical’ (PT) terminology since the late 1990s and the practice of maintaining tropical-type alerts since 2005, even if a strong storm is in the latter stages of completing ET. The application of common terminology and alerting practices over broader (international) scales will gradually become common knowledge with end-users. This is an extremely important component of building awareness and eliciting the desired response from the citizens forecast to be affected by the storm. There remains a significant challenge in our ability to convey to the public and stakeholders that the ET process should not necessarily imply a diminished or “downgraded” threat. It is unfortunately still very common for the Press and even broadcast meteorologists to deliver statements such as “…the storm has been downgraded to post-tropical…”. A large part of the issue is with our naming and storm classification convention, and perhaps a change is in order to implement basic categorization levels for PT cyclones. Currently, in North America, a PT storm may be producing maximum winds equivalent to a strong category-one hurricane (e.g. 75 or 80 knots), or simply gale-force winds of 35 knots. Perhaps PT storms carrying winds of 65 kts or more ought to be called “Post-tropical Hurricane …”? This classification is however, somewhat self-contradictory. More graphical forecast products that focus on the most impactful parameters during ET are necessary. At the CHC there are efforts underway to exploit new production software that would facilitate the construction of maps showing the swaths of heavy rainfall, high winds, storm surge and waves from a TC undergoing ET (Figure 18). Since the transition process involves drastic changes in the storm structure, these products are necessary to draw focus away from the actual storm track line, and highlight the areas most likely to be impacted. Below is a prototype graphic showing the pattern of track-relative wind and rainfall with a hypothetical storm undergoing ET. A similar map for ocean-related impacts could also be established. These additional products would accompany the traditional track forecast maps that often contain a track error cone to reflect unknowns about the track. (Fogarty)

Figure 18. Example for new warning product during ET

20

4.2.5 Processes during ET 4.2.5.1 Inner Core Structure and Intensity Change To capture changes to the inner-core structure of the TC during ET, NOAA has recently conducted missions with its P3 (Aberson et al. 2006) and G-IV (Aberson 2010) aircraft during the time preceding and during the extratropical transitions of Hurricanes Earl (2010), Sandy (2012), and Arthur (2014). The observations include those at flight level, surface wind speeds from the Stepped Frequency Microwave Radiometer, dropwindsondes, and from the Doppler radars aboard all three aircraft. As Arthur developed, its centre was tilted to the south with altitude, but as Arthur intensified into a hurricane, the tilt lessened (Figure 20). By the time of the first dedicated ET flight with the G-IV on the evening of 04 July, the low-level centre was becoming elongated to the northwest, and Arthur had again become asymmetric, with strongest winds on the east side, and the radius of maximum wind speed much more distant from the centre than previously. During the final P3 flight about 12 h later, the convection had rapidly weakened (as shown by the limited Doppler coverage in mid levels), but a very strong low-level jet is still seen offshore the Maine Coast. The aircraft was too close to shore to capture changes to the precipitation structure of the storm. The Hurricane Ensemble Data Assimilation System (HEDAS) has been developed to assimilate aircraft and other data into the operational Hurricane Weather Research and Forecast Model and has shown promise in providing accurate analyses and improved forecasts (Aksoy et al. 2013). Unlike the Doppler radar wind analyses (Figure 20), these provide three-dimensional pictures of the tropical cyclone kinematic and thermodynamic structure even where no data exist. Figure 21 shows the evolution of the temperature and wind fields at four levels during the final six P3 flights into Hurricane Earl. The warm core is maintained best in middle levels and weakens most rapidly near the tropopause. The radius of maximum wind speed also increases with time as Earl undergoes ET, and this expansion occurs at all levels concurrently. Hurricane Sandy underwent an aborted ET on 26 October, and a final transition just before landfall on 30 October. During the first episode, the surface air and dewpoint temperatures derived from HEDAS (Figure 22) remained warm, suggesting that cold water was not a factor in this transition, and also the possibility that Sandy would regain tropical characteristics. During this time, though, the convection disappeared around the storm as shown by the time series of convective available potential temperature and total precipitable water. These values soon increased, allowing Sandy to again become a TC. During the first, aborted transition, the wind field expanded as in a standard transition, leading to the ultimate landfall of a very large system in the U. S. northeast. (Aberson)

21

Figure 19. Composites of Airborne Doppler wind analyses and dropwind sonde data at 7 and 0.5 km altitude from NOAA aircraft for Hurricane Arthur. Best track for Arthur from NHC.

22

Figure 20. HEDAS temperature and wind velocity analyses at 200 hPa (top), 500 hPa (second row), 700 hPa (third row), and 850 hPa (bottom) expressed as a perturbation from the

mean in the full domain, for Hurricane Earl

Figure 21. HEDAS-derived time series of surface air and dew point temperatures (left), convective available potential temperature (middle), and total precipitable water (right), along with best track intensity for Hurricane Sandy. The arrows

along the top represent the times when airborne Doppler radar data are assimilated into the analyses. As aircraft reconnaissance missions are not flown into TCs during ET except for cases in which the TC poses a threat to land within the northern Atlantic basin (like in the cases in Section 4.2.5.1), direct observations of TC intensity during ET are virtually non-existent. While scatterometers offer promise for remotely assessing TC intensity during ET, their surface wind retrievals are ambiguous in regions of intense precipitation and present-day observation capabilities permit only one overpass of a TC every twelve hours at best. Consequently, intensity estimates obtained using the subjective Dvorak Technique (DT; Dvorak 1984) and/or objective Advanced Dvorak Technique (ADT; Olander and Velden 2007) form the basis for the majority of both real-time and historical intensity estimates for TCs during ET (J. Beven 2012, personal communication).

23

While the DT and ADT perform well for purely tropical cyclones (e.g., Brown and Franklin 2004, Olander and Velden 2007), DT- and ADT-derived TC intensity estimates become less reliable during ET (Velden et al. 2006). This degradation in reliability arises primarily because the empirical relationships between cloud patterns and TC intensity that underlie the DT and ADT do not hold during ET, during which the interaction of the TC with an upstream baroclinic zone causes the TC’s cloud field to become highly asymmetric (e.g., Klein et al. 2000) while providing an energy source to partially maintain the intensity of the transitioning TC’s. To illustrate the degradation of TC intensity estimates during ET, we conduct a verification of ADT-derived intensity estimates for twelve near-land north Atlantic basin TCs between 2005 and 2008 that are within two days of completing ET. In the composite mean (n = 18), ADT-derived intensity estimates during ET exhibit 8.47 hPa and 13.12 kt weak biases for minimum sea-level pressure and maximum-sustained 10-m wind speed, respectively. While DT- and ADT- intensity estimates are known to be less reliable and appear to be weak-biased during ET, the precise magnitude and temporal evolution of such bias remain unknown owing to a lack of observations against which verification may be conducted. As a first step toward the development of a means through which improved satellite-derived TC intensity estimates may be obtained, we utilize a novel method based upon numerical simulations and synthetic satellite imagery to evaluate ADT-derived TC intensity estimates during ET. Observing system simulations of five representative north Atlantic TCs – Edouard (1996), Erin (2001), Noel (2007), Ophelia (2011), Leslie (2012) – that completed ET over water are conducted using version 3.4.1 of the Advanced Research Weather Research and Forecasting (WRF-ARW; Skamarock et al. 2008) model at convection-permitting horizontal grid spacing (∆x = 4 km). A proxy for the observed minimum sea-level pressure and maximum-sustained 10-m wind is obtained from hourly model output, whereas a proxy for observed longwave infrared satellite imagery is obtained from synthetic satellite imagery derived from numerical model output (e.g., Bikos et al. 2012, Grasso et al. 2014). As synthetic satellite imagery retrievals are highly sensitive to numerically-simulated cloud ice properties (Grasso and Greenwood 2004), five simulations of each case are conducted, one each utilizing the WSM6, WDM6, Thompson, Morrison, and Milbrandt-Yau microphysical parameterizations. Each observing system simulation is found to produce realistic synthetic cloud structures prior, during, and after ET, at least in a qualitative sense. However, as each microphysical parameterization differs in how cloud ice is represented, there is substantial variability in the qualitative and quantitative characteristics of these structures for each case. Error statistics are normalized to the ET timeline of Evans and Hart (2003) to permit comparison between cases. For minimum sea-level pressure, the composite mean (n = 25) ADT-derived intensity is weak-biased by 10 hPa one day prior to the start of ET. This weak bias grows steadily to 23 hPa at the mid-point of ET, levelling off thereafter. For maximum-sustained 10-m wind, the composite mean ADT-derived intensity is weak biased by 10 kt one day prior to the start of ET. This weak bias grows to 25 kt at the start of ET before decreasing to 16 kt at the end of ET. The erosion of deep, moist convection within the ADT evaluation radius of 136 km from the TC centre (Olander and Velden 2013) that occurs during ET is found to result in a rapid change of ADT scene type from eye, central dense overcast, or embedded centre (case-dependent) to curved band or shear, thereby resulting in a rapid decay in ADT-derived intensity estimates for both intensity metrics. Further research is necessary, however, to quantify the error bars that should be applied to these bias estimates and more fully understand the precise causes of the degradation in ADT-derived intensity estimates (and how such degradation may be mitigated) during ET. (Clark Evans) 4.2.5.2 Interaction with midlatitudes A general characteristic of the direct impact of ET is the gradual amplification of an upper-level ridge and the gradual (formation and) acceleration of a midlatitude upper-level jet streak

24

directly downstream of the ET system along with the retard of the eastward propagation of a potentially pre-existing Rossby wave (RW) packet (Keller and Grams, 2014). This modification of the upper-level midlatitude flow is the result of diabatically enhanced net transport of low-PV air from low levels to the tropopause and the isentropic advection of the low-PV air to the jet by the upper-level diabatic outflow of the TC undergoing ET (Figure 22, Grams et al., 2013a, Archambault et al. 2013, Grams et al., 2011). By advecting tropical, warm, moist air poleward into the extratropics a TC can initiate midlatitude impact during its tropical stage. When this tropical air impinges on a baroclinic zone or experiences upper-level forcing a diabatic Rossby wave (DRW)-like cyclone (see Jangmi (2008) or Choi-Wan (2009); Grams et al. 2013a,b; Keller and Grams, 2014) or a predecessor rain event (PRE) can be generated (Grams and Archambault, 2014). Before the actual TC/ET outflow interacts with the midlatitude flow, the diabatic outflow of this DRW or PRE can initially build a ridge and accelerate the jet similar to the ET system itself. In essence, the diabatic outflows associated with these weather systems (ET,-WCB, DRW, or PRE) all result in a substantial ridge-building and acceleration of the upper-level midlatitude jet directly downstream of the ET system, and a retard of the eastward propagation of a potentially pre-existing Rossby wave packet. (Grams)

Figure 22. Figure 5 from Grams et al. (2013a). 3-D view of ET Jangmi at 00 UTC 30 Sep 2008 (left) and 12 UTC 01 Oct Sep 2008 (right). 1.5 PVU PV surface (blue shading),320 K surface of Θe (transparent grey shading), representing 3-D baroclinic

zone. |v| = 60m s−1 (green shading), highlighting upper-level midlatitude jet. Θ at 990 hPa (shading at bottom, brown colours >300 K, green ≈290 K, blue <280 K) and geopotential at 990 hPa (black contours, every 250 gpm). Paths of

representative trajectories (every 100th, starting 28/12UTC → ending 30/12UTC (left) and 30/12UTC → 02/12UTC (right)) coloured by PV of air parcel moving along trajectory. Low-PV air (PV<0.6 PVU) grey shades, higher PV values (PV>0.6 PVU)

in red shades (legend bottom right).

Few case studies of individual ET events in the SWIO have ever been published, with Griffin and Bosart (2014) analyzing TC Edisoana (1990), the deepest post-ET reintensification noted in SWIO records. As this is one of the few case studies of ET noted, it is difficult to state how representative the findings of this one study are with respect to the population of ET cases in the SWIO. TC Edisoana formed east of Madagascar near 70°E, drifted westward to near 60°E, turned poleward and underwent ET between 30-35°S on 7-8 March, and deepened ~40 hPa (to under 940 hPa) in the 24 h ending 1200 UTC 9 March as an EC as it passed just east of the Kerguelen Islands. As demonstrated by a Hart phase space analysis (not shown), EC Edisoana behaved similarly to other North Atlantic TCs that deepened explosively as ECs after ET (Hart 2003; Hart et al. 2006). EC Edisoana deepened rapidly as the half-wavelength downstream of the polar trough collapsed around 1200 UTC 8 March when the storm was located in the favourable equatorward entrance region of the downstream jet streak (Figure 23a). During the EC rapid intensification (RI) period, positive differential potential vorticity (PV) advection by the irrotational wind (~15-20 PVU

25

day-1; not shown) eroded the poleward portion of the polar trough and allowed the downstream jet to backbuild, helping EC Edisoana to maintain its position in the equatorward entrance region of this jet. Simultaneously, negative PV advection by the non-divergent wind in the base of the polar trough (not shown) served to bring high-magnitude PV air east faster than the high-magnitude PV air to the south. The ensuing evolution is reminiscent of the formation of a PV hook during the LC2 lifecycle discussed in Thorncroft et al. (1993) and Thorncroft and Jones (2000). The interaction between EC Edisoana and the midlatitude flow served to separate the merged subtropical and polar front jets into two distinct pieces upstream and downstream of the cyclone. This separation process was apparent around 0000 UTC 9 March as the strong downstream merged jet spanning the crest of the phased ridge separated from a weaker, upstream merged jet streak in the base of the polar trough (Figure 23b). A simultaneous increase in the maximum wind speed observed in the upstream jet streak (~ 20 m s-1 in 6 h) further enhanced the favourable synoptic-scale forcing from the associated ageostrophic circulations over EC Edisoana, as the cyclone was clearly located in both the equatorward entrance region of the downstream jet and the poleward exit region of the upstream jet.

Figure 23. Plot of 250 hPa isotachs (shaded per colour bar; m s-1), 1000-500 hPa thickness (dashed contours; dam; 540 and below in blue), 500 hPa upward vertical motion (purple contours every 5 µb s-1), and mean sea level pressure (black

contours every 4 hPa) for (a) 1200 UTC 8 March, (b) 0000 UTC 9 March, and (c) 1200 UTC 9 March 1990. The superposition of synoptic-scale forcing for ascent as described above can often lead to rapidly deepening cyclones. Unlike the paradigm of rapidly-deepening midlatitude cyclones with ascent in both the equatorward entrance region of a downstream jet and the poleward exit region of an upstream jet (e.g. Uccellini and Kocin 1987), EC Edisoana begins deepening in the presence of only a single midlatitude jet (Figure 25a) then develops a secondary jet streak on its equatorward side around the base of the upstream trough such that the surface cyclone associated with EC Edisoana is in the poleward exit region of this new secondary upper level jet streak (Figures. 23 b-c). This development of the secondary jet streak upstream of EC Edisoana can be associated with the approach of the polar trough such that the surface cyclone “phase locks” with the polar trough (Hoskins et al. 1985). This phase locking allows for a mutual intensification of

26

both the surface and upper-level PV anomalies and the intensification of the upstream jet streak in the base of the polar trough. The upstream jet streak is positioned so that its poleward exit region is superposed with the equatorward entrance region of the downstream jet so that strong synoptic-scale forcing for ascent can be inferred over EC Edisoana (Figure 23b). As the upstream jet streak intensifies, the downstream jet streak weakens and progresses further away from EC Edisoana near the end of the RI phase (Figure 23c), leaving the EC with forcing for ascent from only the poleward exit region of the upstream jet streak. (Bosart) The interaction between a TC and the strong climatological jet in the northwest Pacific is well described by Archambault et al. (2012). However, recurving TCs in the northwest Pacific typically encounter only one jet stream, not two as noted by Griffin and Bosart (2014). Given the dual jet streams that are also present in south Pacific ET cases (Sinclair 2002, 2004), more complicated TC-jet interactions may be more common in Southern Hemisphere TC basins. Further studies of both individual cases and the population of ET events would likely provide additional insights into the nature and potential complexity of the interactions between TCs and the jet stream in the SWIO and elsewhere. (Griffin) It is generally believed that roughly 20-30% of tropical disturbances with characteristics favourable for development actually form TCs. Approximately one third of western Pacific tropical cyclones undergo ET. The occurrence of ET implies that TCs reach, and influence, the midlatitude jet and accentuate the severity of downstream weather systems in many cases.

Figure 24. (a) and (b) Relative vorticity (10-5/s) and streamlines and 850 hPa. ‘C1’ refers to cyclone 1, TCS025 refers to the remnants of TCS025; (c) Temperature (K) at 850 hPa. Arrows denote orientation of strongest flow.

27

Unknown is the ultimate fate of the 70-80% of tropical disturbances that do not develop into TCs. Do they simply decay in the tropics? Or do some produce important weather phenomena far removed in space and time from their period of scrutiny as candidate tropical cyclones? This question was investigated by Davis et al. (2013), in which three cyclones were examined that intensified into significant weather systems poleward of the tropics, yet whose roots of vorticity and moisture were in the deep tropics. The study made extensive use of the full suite of model-based products from the Integrated Forecast System (ECMWF) including model tendency fields as part of the additional diagnostic suites produced during the Year of Tropical Convection (YOTC). The study also combined satellite data with observations collected from the DLR Falcon. Three cases were examined. Two of these cases had been tracked, and flown, as part of the Tropical Cyclone Structure 2008 (TCS08) component of the THORPEX Pacific Asia Regional Campaign (TPARC). The TCS08 project sought to document genesis processes and structure of tropical cyclones. Two candidates for TC formation were TCS025 and TCS037. Neither became a TC, however, the remnants of both systems produced mesoscale cyclones with gale-force winds near the coast of Japan.

Figure 25. Cross section (soundings numbered 7-12) of equivalent potential temperature (5 K contour interval, values of 350 K or above in red, 325 K or below in blue), wind barbs, and water vapour mixing ratio (g/kg, shaded ) from the water

vapour differential absorption lidar (DIAL) aboard the DLR Falcon (g/kg) valid at approximately 00 UTC 30 August The combination of model-derived fields and direct observations indicated that an important component of cyclogenesis was the organization of deep, moist convection through baroclinic processes, that is, organized vertical motion that resulted as vorticity in the lower troposphere encountered increasing baroclinicity at higher latitudes. In Figure 24 is a summary of a mesoscale cyclone that developed on 30-31 August 2008. The cyclone (marked ‘C1’) had its roots in the circulation and moisture of TCS025, but as seen in Figure 24, is a development that is clearly distinct from the original vorticity maximum of TCS025. Development occurs as the moist southeasterly flow (Figure 25) encounters weak baroclinicity near 35°N associated in part with the gradient of sea-surface temperature. Weak frontal lifting was believed sufficient to initiate widespread convection. The cross section (Figure 25) normal to the southeast flow, shows the

28

high values of equivalent potential temperature confined within a narrow zone and extending above the boundary layer. The cyclone development in this, and in the other two cases examined, was not classified as ET because a TC did not precede the development. Nonetheless, the development of frontal features and the formation of a highly asymmetric wind field bore resemblance to the process of ET. The life cycle of these cyclones was short, on the order of 2 days. There was no predictive skill of any of these cyclones in the ECMWF ensemble beyond 3 days. (Davis)

4.2.5.3 Downstream Impact Relocation experiments for T-PARC Typhoon Jangmi (2008) confirmed the existence of a critical bifurcation point in the upper-level midlatitude flow, that governs the downstream impact of a real ET (Figure 26; Grams et al. 2013b). Such bifurcation points are located on the axis of an (initially weak) upstream trough (firstly shown in an idealised setup by Scheck et al. 2011) and primarily explain the reduced predictability for the track of the ET system and the downstream flow evolution. If the ET system tracks to the east and poleward of such a bifurcation point it reaches a zone with strong extratropical upper-level forcing, the ET system strongly reintensifies and subsequently a pronounced downstream RWT is triggered. Otherwise the ET system decays and a much more zonal flow occurs downstream. In addition, Riemer and Jones (2013) showed the existance of a second bifurcation point on the axis of the downstream ridge that allows to distinguish the Northeast and Northwest pattern during ET, identified by Harr et al. (2000).

Figure 26. Figure 9 from Grams et al. (2013b). Trough-relative stream function (shaded every 0.5×106 m2/s) vertically averaged from 100 to 300 hPa and associated trough-relative non-divergent wind vectors (scale for wind speed in m/s below the colour bar). Contours: ur + vr = 0, 10 m/s (white solid). Time shown is 0600 UTC 1 October 2008. Tracks are

shown relative to the estimated zonal trough propagation speed of 0.5° /h and centred on the TC position (open square) at the time shown. A geographical coordinate system centred on the bifurcation point is used.

In a more generalised setup, simulations initialised from composite fields (Grams and Archambault, 2014) confirmed that Western North Pacific ET retards the eastward propagation of RW packets, strongly amplifies the downstream Rossby wave guide, triggers the downstream development of ridge-trough couplets in the Eastern North Pacific and over North America and thus the evolution of a RWT. This downstream development alters the intensity and location of HIW in regions remote from the ET system. (Grams)

29

By analysing the eddy kinetic energy budget in several forecast scenarios, extracted from the ECMWF EPS for the ET of Choi-Wan (2009) and Hanna (2008), Keller et al. (2014) assessed the role of the two storms in the modification of the donwstream midlatitude flow. For Typhoon Choi-Wan, the impact of the transitioning TC on the amplification of the downstream wave train depended strongly on the phasing between the storm and a preexisting midlatitude wave train. In addition, the interaction with Choi-Wan and a frontal wave ahead of the storm was crucial for the outcome. Further downstream, the ET of Typhoon Choi-Wan led to an autumn heat wave in the Vancouver region and a cold-air outbreak east of the Rocky Mountains (Keller and Grams, 2014). For the ET of Hurricane Hanna, the duration of the baroclinic conversion during the interaction with the midlatitude flow was crucial. If baroclinic conversion continued until Hanna approached a weakly amplified midlatitude trough, the two systems interacted, what resulted in a reintensification of Hanna and a further amplification of the trough. In case the baroclinic conversion ceased to exist prior to the interaction with the trough, no reintensification/amplification occurred. These findings could be further confirmed by conducting ensemble sensitivity experiments (Keller, 2014). (Keller) The importance of the upstream midlatitude flow configuration to the ET and reintensification of Hurricane Earl (1998) was demonstrated by McTaggart-Cowan et al. (2001): the system failed to redevelop when the upshear trough was removed from numerical integrations immediately prior to ET. Using a singular vector analysis, Corbosiero et al. (2012) show that the bulk of sensitivity lies in the amplitude and phasing of the ridges the flank the midlatitude trough. A stronger downstream ridge enhances Earl's final post-transition intensity, while a more progressive upstream ridge reduces it. The latter is of particular interest because of the presence of Tropical Storm Isis in the eastern North Pacific, a feature whose presence may have affected the anticyclonic Rossby wave break over the North American continent that formed the upstream ridge.

Figure 27. Tracks of Hurricane Earl in 5-day integrations starting at 1200 UTC 1 September (left panel, colour-coded for jet wind speeds as shown in the right panel legend). The potential temperature on the dynamic tropopause (2 PVU surface) at 1200 UTC 6 September are shaded according to the colour bar. Initial wind speeds in the jet and their relation to Earl's final

intensity are shown in the right panel. Through a set of vortex insertion tests, Corbosiero et al. (2012) found that Isis exerted relatively little influence on the breaking wave, and thus on the ET and reintensification of Earl in the western North Atlantic. Instead, the investigators found that the strength of the zonally extended North Pacific jet approximately five days before Earl's transition had a dramatic impact on both the anticyclonic wave break (upstream ridge) and the storm's final intensity as shown in the Figure 27. The observed 150 kt jet is found to have been close to a “sweet spot” for reintensification, with wind speeds more than 20 kt from this value resulting in dramatic reductions in Earl's strength following ET. The study concludes that in this case, the structure of the midlatitude flow far upstream of the storm's recurvature point has a more meaningful impact on the

30

ET process than does the presence of a tropical cyclone interacting with the subtropical jet within one Rossby radius of the transitioning TC. (McTaggart-Cowan) The remote impact of North Atlantic ET on the predictability of Mediterranean severe weather, which is often induced by Rossby wave breaking and mostly occurs in the same season as ET, was investigated in numerical models with different methods: 1. Sensitivity to the representation of convection (Pantillon et al. 2013a). The track of Hurricane Helene (2006) was poorly forecast by operational models during ET. Here a convection-permitting model run over a large domain improved the track forecast compared to a lower-resolution run. The higher resolution improved the phasing of Helene with a Rossby wave train, which in turn improved the track of Helene. The improved phasing was allowed by differences along the Rossby wave train upstream of Helene rather than in the convection of the hurricane. 2. Sensitivity to the initial conditions (Pantillon et al. 2103b). The ET of Hurricane Helene (2006) was further investigated in the operational ECMWF ensemble forecast. Clustering the ensemble members revealed two scenarios that linked the ET of Helene with the development of a tropical-like cyclone over the Mediterranean (Medicane). Only the correct phasing of Helene with a Rossby wave train allowed the correct track of Helene. The correct phasing also increased the probability of developing a Medicane in the ensemble forecast (Figure 28). 3. Sensitivity to the presence of tropical cyclones (Pantillon et al, 2014). The remote impact of tropical cyclones on downstream wave breaking was investigated during HyMeX (hydrological cycle in the Mediterranean experiment). Hurricanes Leslie, Rafael and Sandy were filtered out in numerical experiments, which were compared to control model runs. The tropical cyclones weakly impacted or even decreased the forecast precipitation over the Mediterranean. Their interaction with the midlatitude flow modified but did not trigger downstream wave breaking. (Pantillon) A study by Archambault et al. (2014b) uses ensemble prediction system data to illustrate dynamical links between the recurvature and ET of western North Pacific TC Malakas in late September 2010 and increased model forecast error and uncertainty downstream. Operational GFS forecasts fail to predict the recurvature of Malakas and as a result suffer from poor short-to-medium range skill on a near-hemispheric scale over several days. A parallel is found between the operational GFS forecasts and a numerical simulation in which Malakas is removed from model initial conditions using PV surgery: In both the operational forecasts and the “no TC” simulation, the extratropical flow is much less amplified downstream than observed. Based on an analysis of members of the ECMWF ensemble prediction system, forecasts of the timing and strength of Rossby wave amplification and dispersion in late September 2010 are highly sensitive to the forecast track of TC Malakas and the forecast strength of the objectively defined TC–extratropical flow interaction associated with TC Malakas. (Archambault)

31

Figure 28. Link between Hurricane Helene (2006) and a Mediterranean tropical-like cyclone (Medicane) in the operational ECMWF ensemble forecast, partitioned in 2 clusters. Tracks of Helene until 1200 UTC 26 September 2006 and tracks of the Medicane on 26 September 2006; tracks in the ECMWF analysis (solid black line), deterministic forecasts (dashed

black lines), and ensemble forecasts (shading) initialized at 0000 UTC from 20 to 23 September 2006. Ensemble members that predict the Medicane are in colour (red and blue of 22 and 23 September 2006, respectively).

Dots indicate the position of Helene at 0000 UTC.

4.2.6 Hybrid Systems Mediterranean Sea is from time to time affected by very intense tropical-like cyclonic storms. Although the frequency is low, these small scale systems present a significant challenge for the forecasters and can produce important life and property damages. We can consider the tropical-like Mediterranean cyclones as hybrid systems. During the initial stages midlatitude baroclinic processes used to play the key role, in association with a cut-off low at upper levels (Emanuel, 2005). But, once developed, they are maintained in the same way as tropical cyclones, by inducing large enthalpy fluxes from the sea (Pytharoulis et al., 2000). In the mature stage they present a strong pressure gradient and a low-level warm core.

The cyclones initiate and develop over the sea, in an area of deep convection usually organised as convective cloud bands spiralling into a minimum low pressure. In some cases they have a clear eye surrounded by a convective cloud pattern, looking like a “small hurricane” in satellites images. This is the reason for using the word Medicanes (acronym for Mediterranean hurricanes).

32

Figure 29. MODIS image of the tropical-like Mediterranean cyclone on 19th November 2013 at 15:30 UTC.

The centre of the cyclone is just off the Sardinia coast.

Figure 30. Cyclone phase space diagram from UKMET model. It shows the evolution of the Mediterranean storm between

18th November 00UTC and 21st November 00UTC. Source: Florida State University

One of these cyclones developed on 18th-19th November 2013 over the western Mediterranean Sea, between the Balearic Islands and Sardinia. The strong and slow moving hybrid storm produced intense rains over Sardinia with maxima in 24-hour as higher as 467 mm (18.39"), killing at least eighteen people.

Although sea waters were only relatively warm, around 20°C, the system developed in an area of large air-sea thermodynamic disequilibrium due to the cold air associated to the cut-off. Finally the storm was able to develop a shallow low-level warm core, as can be seen in the cyclone phase space diagram, while the low deepened to a central pressure near 990 mb. In the diagram is noticeable the period with a warm core, between 18th and 19th November, evolving toward an asymmetric cold core later in its life cycle, a fairly typical development. (Elizaga)

33

4.2.7 Summary and outlook on pathways for further research This report summarizes the advances in understanding ET that have been made since the IWTC-VII in 2010. Owing to a diversity of Working Group members from different countries, institutions and scientific background, it describes a diversity of approaches and results in further understanding the processes involved in the ET of TCs and the associated downstream impact, as well as covering the current knowledge on climatological distributions of ET in various ocean basins. In terms of objectively classifying ET, new methods focus on differences in the ET environment, the CPS pathway and the isentropic PV on Theta surfaces as a discriminator for ET. Using novel reanalysis data sets, the researchers have derived new climatologies on ET in distinct ocean basins, particularly the southwestern Indian Ocean and the eastern North Pacific Ocean. North Atlantic Ocean climatologies are now extended to cover the typical impact of interactions between a TC and an extratropical cyclone on the TC track. Some further advances are also made with respect to TCs and their potential ETs in the South Atlantic Ocean. Recent studies also dealt with climatologically investigating the (downstream) impact of ET, e.g. in terms of typical precipitation patterns or the amplification of midlatitude Rossby wave trains. Advancements could also be achieved with respect to further developing ET forecast strategies at the JTWC, the communication during ET events and the potential hazard they may pose to the public and the value of targeted observations on ET forecasts. First attempts are also made on investigating the general reduction in predictability during ET. New developments in assimilating aircraft reconnaissance data into the Hurricane Weather and Research Forecast model are shown result in improved analysis and forecasts for the inner core structure during ET. An attempt is made to estimate the bias in intensity estimates derived through the Dvorak or Advanced Dvorak Technique during ET by using synthetic satellite images from high-resolution model runs. Case studies on ET events in different ocean basins, as well as on the remnants of tropical disturbances that did not develop into a TC, help to further reveal the important processes. Among others, these comprise the advection of PV by the nondivergent wind, the diabatic PV modifications, the warm-conveyor-belt (WCB) like ascend along the baroclinic zone or the source of additional eddy kinetic energy. The downstream impact of an ET event depends on the phasing between the transitioning TC and a preexisting midlatitude wave train and may be even sensitive to the flow configuration far upstream of the recurvature point. Idealized as well as real case studies point to the existence of a bifurcation point in the storm-relative stream function which determines whether the storm moves ahead of the trough and thus into a favourable position for reintensification. The associated impact on the midlatitudes could be shown to play a significant role in the development of high impact weather in the Mediterranean or over North America. The new advances in understanding ET also suggest areas that need further research. This holds for developing an objective classification of ET, a prerequisite for further standardizing climatological studies and intensity estimations. More systematic studies are needed on the physical processes determining the relative position of the ET system and midlatitude bifurcation points and on how well these are represented in current NWP models. Further, more insight is needed on how a TC potentially alters the midlatitude environment in a way favourable for extratropical reintensification prior to ET (via direct impact of TC WCB-like outflow, or DRW or PREs ahead of the TC?). Schafler and Harnisch (2014) showed that current NWP substantially

34

misrepresents the atmospheric moisture in the inflow region of a WCB associated with ET and resulting in forecast error for the ET process. Therefore, future field campaigns would be helpful to better understand the role of condensational heating and its representation in numerical models during ET. References Anwender, D.,C. Cardinali, S. C. Jones, S.C., 2014: Data denial experiments for extratropical

transition, Tellus, 64, 19151 Aberson, S. D., 2010: 10 years of hurricane synoptic surveillance (1997–2006). Mon. Wea. Rev.,

138, 1536–1549. Aberson, S. D., M. L. Black, R. A. Black, J. J. Cione, C. W. Landsea, F. D. Marks Jr., and R. W.

Burpee, 2006: Thirty years of tropical cyclone research with the NOAA P-3 aircraft. Bull. Amer. Meteor. Soc., 87, 1039–1055.

Aksoy, A. S. D. Aberson, T. Vukicevic, K. J. Sellwood, S. Lorsolo, and X. Zhang, 2013: Assimilation of high-Resolution tropical cyclone observations with an ensemble Kalman filter using NOAA/AOML/HRD’s HEDAS: Evaluation of the 2008–11 vortex-scale analyses. Mon. Wea. Rev., 141, 1842–1865.

Archambault, H. M., L. F. Bosart, D. Keyser, and J. M. Cordeira, 2013: A climatological analysis of the extratropical flow response to recurving western North Pacific tropical cyclones. Mon. Wea. Rev., 141, 2325–2346.

Archambault, H. M., D. Keyser, L. F. Bosart, C. A. Davis, and J. M. Cordeira, 2014a: A climatological analysis of the extratropical flow response to recurving western North Pacific tropical cyclones. Mon. Wea. Rev., in review.

Archambault, H. M., P. A. Harr, and R. W. Moore, 2014b: A diagnosis of the poorly forecast interaction between recurving tropical cyclone Malakas (2010) and the extratropical flow. 31st Conference on Hurricanes and Tropical Meteorology, San Diego, CA.

Bikos, D., and coauthors, 2012: Synthetic satellite imagery for real-time high resolution model evaluation. Wea. Forecasting, 27, 784–795.

CIRA-RAMMB, cited 2014: Tropical Cyclone Shear/Instability Phase Space Study. [Available online at: http://rammb.cira.colostate.edu/research/tropical_cyclones/tc_phase_space/.]

Corbosiero, K., R. McTaggart-Cowan and L. F. Bosart, 2012: The impact of recurving eastern Pacific tropical cyclones on the downstream North American flow pattern. 4th International Workshop on Extratropical Transition. May 2012, Mt. Gabriel, Quebec.

Dias Pinto, J. R., and R. P. da Rocha (2011), The energy cycle and structural evolution of cyclones over southeastern South America in three case studies, J. Geophys. Res., 116, D14112, doi:10.1029/2011JD016217.

Dias Pinto, J. R., M. S. Reboita, and R. P. da Rocha (2013), Synoptic and dynamical analysis of subtropical cyclone Anita (2010) and its potential for tropical transition over the South Atlantic Ocean,J. Geophys. Res. Atmos., 118, 10,870–10,883, doi:10.1002/jgrd.50830.

Davis, C. A., S. C. Jones, D. Anwender, L. Scheck, and J. Badey, 2013: Mesoscale cyclogenesis over the western North Pacific Ocean during TPARC. Tellus A, 65, 18621.

Dvorak, V. F., 1984: Tropical cyclone intensity analysis using satellite data. NOAA Tech. Report NESDIS 11, 47 pp.

Emanuel, K. A.: Genesis and maintenance of “Mediterranean hurricanes”, Adv. in Geos., 2, 217–220, 2005.

35

Evans, J. L., and R. E. Hart, 2003: Objective indicators of the life cycle evolution of extratropical transition for Atlantic tropical cyclones. Mon. Wea. Rev., 131, 909-925.

Felker, S. R., B. LaCasse, J. S. Tyo, and E. A. Ritchie, 2011: Forecasting Post-Extratropical Transition Outcomes for Tropical Cyclones Using Support Vector Machine Classifiers. J. Atmos. Oceanic Tech., 28, 709-719.

Fita, L., R. Romero, L. Luque, K. Emanuel, and C. Ramis (2007), Analysis of the environments of seven Mediterranean tropical-like storms using an axisymmetric, nonhydrostatic, cloud resolving model, Natl. Hazards Earth Syst. Sci., 7, 41–56.

Grams, C.M., Keller. J.H., Lang S.T.K., 2014. The role of extratropical transition in the degradation of numerical weather forecasts for the midlatitude flow. Manuscript in preparation for Geophys. Res. Lett.

Grams, C.M. and H. M. Archambault, 2014. Quantification of the impact of extratropical transition on the midlatitude flow: From case studies to a composite view. Poster presentation during 16th Cyclone Workshop, 26 Sep 2013, Sainte-Adèle, Québec, Canada and in preparation for Mon. Wea. Rev.

Grams C.M., 2013. Quantification of Hurricane Sandy's Impact on the Midlatitude Flow. Presentation during the 16th Cyclone Workshop, 23 Sep 2013, Sainte-Adèle, Québec, Canada

Grams C.M., S. C. Jones, C. A. Davis, P. A. Harr, and M. Weissmann, 2013a. The impact of Typhoon Jangmi (2008) on the midlatitude flow. Part I: Upper-level ridgebuilding and modification of the jet. Q. J. R. Meteorol. Soc., 139, 2148–2164.

Grams C.M., S. C. Jones and C. A. Davis, 2013b. The impact of Typhoon Jangmi (2008) on the midlatitude flow. Part II: Downstream evolution. Q. J. R. Meteorol. Soc., 139, 2165–2180.

Grams, C.M., H. Wernli, M. Böttcher, J. Čampa, U. Corsmeier, S. C. Jones, J. H. Keller, C.-J. Lenz and L. Wiegand, 2011. The key role of diabatic processes in modifying the upper-tropospheric wave guide: a North Atlantic case-study. Q.J.R. Meteorol. Soc., 137, 2174–2193.

Grams, C.M., 2011. Quantification of the downstream impact of extratropical transition for Typhoon Jangmi and other case studies. Dissertation, Karlsruhe Institute of Technology, Karlsruhe, Germany. Available online at: http://www.imk-tro.kit.edu/english/4328_5346.php

Grasso, L., D. T. Lindsey, K.-S. S. Lim, A. Clark, D. Bikos, and S. R. Dembek, 2014: Evaluation of and suggested improvements to the WSM6 microphysics in WRF-ARW using synthetic and observed GOES-13 imagery. Mon. Wea. Rev., in press.

Grasso, L. D., and T. J. Greenwald, 2004: Analysis of 10.7-µm brightness temperatures of a simulated thunderstorm with two-moment microphysics. Mon. Wea. Rev., 132, 815-825.

Griffin, K. S., and L. F. Bosart, 2014: The extratropical transition of Tropical Cyclone Edisoana (1990). Mon. Wea. Rev., 142, 2772–2793.

Harr, P. A. and J. M. Dea, 2009: Downstream Development Associated with the Extratropical Transition of Tropical Cyclones over theWestern North Pacific. Mon. Wea. Rev., 137, 1295–1319.

Harr, P. A., R. L. Elsberry, and T. F. Hogan, 2000: Extratropical Transition of Tropical Cyclones over the Western North Pacific. Part II: The Impact of Midlatitude Circulation Characteristics. Mon. Wea. Rev., 128, 2634– 2653.

Hart, R. E., J. L. Evans, and C. Evans, 2006: Synoptic Composites of the Extratropical Transition Life Cycle of North Atlantic Tropical Cyclones: Factors Determining Posttransition Evolution. Mon. Wea. Rev., 134, 553–578.

Hart, R. E., 2003: A Cyclone Phase Space Derived from Thermal Wind and Thermal Asymmetry. Mon. Wea. Rev., 131, 585–616.

36

Hart, R. E., and J. L. Evans, 2001: A Climatology of the Extratropical Transition of Atlantic Tropical Cyclones. J. Climate, 14, 546–564.

Hoskins, B. J., M. E. McIntyre and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 111, 877–946.

Keller, J.H., S.C. Jones, P.A. Harr, 2014: An Eddy Kinetic Energy View of Physical and Dynamical Processes in Distinct Forecast Scenarios for the Extratropical Transition of Two Tropical Cyclones. Mon. Wea. Rev., 142, 2751–2771.

Keller, J.H. and C. M. Grams, 2014. The Extratropical Transition of Typhoon Choi-Wan (2009) and its role in the formation of North American high impact weather. Poster presentation during World Weather Open Science Conference, 16-21 September 2014, Montrèal, Québec, Canada and in preparation/revision for Mon. Wea. Rev

Keller, J.H., 2014: Investigating the downstream impact of tropical cyclones using ensemble sensitivity analysis, Oral Presentation, 31st Conference on Hurricanes and Tropical Meteorology, San Diego, CA, Amer. Meteor. Soc.

Kitabatake, N., 2010: Impact of synthetic wind retrieval on tropical cyclone structures at the extratropical transition stage in the JRA-25 reanalysis, SOLA, 6, 77-80.

Kitabatake, N., 2011: Climatology of extratropical transition of tropical cyclones in the western North Pacific defined by using cyclone phase space, J. Meteor. Soc. Japan, 89, 309-325.

Klein, P. M., P. A. Harr, R. L. Elsberry, 2000: Extratropical Transition of Western North Pacific Tropical Cyclones: An Overview and Conceptual Model of the Transformation Stage. Wea. Forecasting, 15, 373–395.

Kofron, D. E., E. A. Ritchie, and J. S. Tyo, 2010a: Determination of a consistent time for the extratropical transition of tropical cyclones Part I: Examination of Existing Methods for Finding “ET-time”. Mon. Wea. Rev., 138, 4321-4343.

Kofron, D. E., E. A. Ritchie, and J. S. Tyo, 2010b: Determination of a consistent time for the extratropical transition of tropical cyclones Part II: Potential Vorticity Metrics. Mon. Wea. Rev. 138, 4344-4361.

Kucas, M.E., S.J. Barlow, and R.C. Ballucanag, 2014: Subtropical cyclones: Operational Practices and analysis methods applied at the Joint Typhoon Warning Center. Extended abstracts, 31st Conference on Hurricanes and Tropical Meteorology, San Diego, CA, Amer. Meteor. Soc., 1-8.

McTaggart-Cowan, R., L. F. Bosart, C. A. Davis, E. H. Atallah, J. R. Gyakum, and K. A. Emanuel (2006), Analysis of Hurricane Catarina (2004), Mon. Weather Rev., 134, 3029–3053

McTaggart-Cowan, R., J. R. Gyakum, and M. K. Yau, 2001: Sensitivity Testing of Extratropical Transitions Using Potential Vorticity Inversions to Modify Initial Conditions: Hurricane Earl Case Study. Mon. Wea. Rev., 129, 1617–1636.

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.

Olander, T. L., and C. S. Velden, 2013: ADT – Advanced Dvorak Technique Users’ Guide. [Available online at http://tropic.ssec.wisc.edu/misc/adt/guides/ADTV8.1.4_Guide.pdf].

Olander, T., and C. Velden, 2007: The Advanced Dvorak Technique: continued development of an objective scheme to estimate tropical cyclone intensity using geostationary infrared satellite imagery. Wea. Forecasting, 22, 287–298.

Quinting, J. and S. C. Jones, 2014: The impact of tropical cyclones on midlatitude Rossby wave packets: A climatological perspective. Extended abstracts, 31st Conference on Hurricanes and Tropical Meteorology, San Diego, CA, Amer. Meteor. Soc.

37

Pantillon, F., J.-P. Chaboureau and E. Richard, 2014, Remote impact of North Atlantic hurricanes on the Mediterranean during episodes of intense rainfall in autumn 2012. Q.J.R. Meteorol. Soc. doi: 10.1002/qj.2419

Pantillon, F., J.-P. Chaboureau, C. Lac, and P. Mascart, 2013a, On the role of a Rossby wave train during the extratropical transition of hurricane Helene (2006). Q.J.R. Meteorol. Soc., 139: 370–386. doi: 10.1002/qj.1974

Pantillon F., J.-P. Chaboureau, P. Mascart, and C. Lac, 2013b: Predictability of a Mediterranean Tropical-Like Storm Downstream of the Extratropical Transition of Hurricane Helene (2006). Mon. Wea. Rev., 141, 1943–1962. doi: 10.1175/MWR-D-12-00164.1

Picornell, M. A., J. Campins, and A. Jansà, 2014: Detection and thermal description of medicanes from numerical simulation. Nat. Hazards Earth Syst. Sci., 1059-1070.

Pytharoulis, I., G. C. Craig, and S. P. Ballard, 2000: The hurricane like Mediterranean cyclone of January 1995, Meteorol. Appl., 7, 261–279.

Riemer M., and S. C. Jones S.C., 2013. Interaction of a tropical cyclone with a high-amplitude, midlatitude wave pattern: Waviness analysis, trough deformation and track bifurcation. Q.J.R. Meteorol. Soc., early online view.

Riemer, M. and S. C. Jones, 2010: The downstream impact of tropical cyclones on a developing baroclinic wave in idealized scenarios of extratropical transition Q. J. R. Meteorol. Soc., 136, 617–637.

Ritchie, E. A., K. M. Wood, D. S. Gutzler, and S. R. White, 2011: The Influence of Eastern Pacific Tropical Cyclone Remnants on the Southwestern United States. Mon. Wea. Rev., 139, 192–210.

Schäfler, A., and F. Harnisch, 2014: Impact of the inflow moisture on the evolution of a Warm Conveyor Belt. Q.J.R. Meteorol. Soc., published online, doi: 10.1002/qj.2360

Scheck L., S. C. Jones and M. Juckes, 2011. The resonant interaction of a tropical cyclone and a tropopause front in a barotropic model. Part II: frontal waves. J. Atmos. Sci., 68, 420–429.

Schenkel, B. A., and R. E. Hart, 2012: An examination of tropical cyclone position, intensity, and intensity life cycle within atmospheric reanalysis datasets. J. Climate, 25, 3453–3475.

Sinclair, M. L., 2002: Extratropical Transition of Southwest Pacific Tropical Cyclones. Part I: Climatology and Mean Structure Changes. Mon. Wea. Rev., 130, 590–609.

Skamarock, W. C., and coauthors, 2008: A Description of the Advanced Research WRF Version, 3. NCAR Technical Note NCAR/TN-475+STR, 61 pp. [Available online at http://www.mmm.ucar.edu/wrf/users/docs/arw_v3.pdf].

Thorncroft, C. D., and S. C. Jones, 2000: The Extratropical Transitions of Hurricanes Felix and Iris in 1995. Mon. Wea. Rev., 128, 947–972.

Thorncroft, C. D., Hoskins, B. J., and McIntyre, M. E., 1993: Two paradigms of baroclinic-wave life-cycle behaviour. Quart. J. Roy. Meteor. Soc., 119, 17–55.

Truchelut, R. E., R. E. Hart, and B. Luthman, 2013: Global Identification of Previously Undetected Pre-Satellite-Era Tropical Cyclone Candidates in NOAA/CIRES Twentieth-Century Reanalysis Data. J. Appl. Meteor. Climatol., 52, 2243–2259.

Truchelut, R. E., and R. E. Hart, 2011: Quantifying the possible existence of undocumented Atlantic warm-core cyclones in NOAA/CIRES 20th Century Reanalysis data. Geophys. Res. Lett., 38, L08811.

Uccellini, L. W., and P. J. Kocin, 1987: An examination of vertical circulations associated with heavy snow events along the East Coast of the United States. Wea. Forecasting, 2, 289-308.

Valadon, Y., 1992: Variability of Tropical Cyclones over the Southwest Indian Ocean. J. Afr. Meteor. Soc., 1, 45–51.

38

Velden, C., and coauthors, 2006: The Dvorak tropical cyclone intensity estimation technique: a satellite-based method that has endured for over 30 years. Bull. Amer. Meteor. Soc., 87, 1195–1210.

Wood, K. M., and E. A. Ritchie, 2014a: A 40-y climatology of extratropical transition in the eastern North Pacific. J. Climate, 24, 5999-6015.

Wood, K. M. and E. A. Ritchie, 2014b: A 32-year reanalysis intercomparison of tropical cyclone structure in the eastern North Pacific and North Atlantic. Extended abstract presented at: 31st Conference on Hurricanes and Tropical Meteorology; 2014 March 31 - April 4; San

Diego,California. Wood, K. M., and E. A. Ritchie, 2013: An Updated Climatology of Tropical Cyclone Impacts on the

Southwestern United States. Mon. Wea. Rev., 141, 4322–4336. Wood, K. M., and E. A. Ritchie, 2012: The unique behavior and precipitation pattern associated

with Tropical Storm Ignacio (1997). Mon. Wea. Rev., 140, 3347–3360.

______