A two-stage ET classification based on Klein et al. (2000). The onset and completion times correspond to the definitions of Evans and Hart (2003). The “tropical” and “extratropical” labels indicate approximately how the system would be regarded by an operational forecast center. Figure reproduced from Jones et al. (2003, their Fig. 11).
Tracks of TCs that completed the transformation stage of ET for the (a) NATL [1981–2010; ET designations from HURDAT2 best track data, described in Landsea and Franklin (2013)], (b) WNP (1981–2010; ET designations from Japan Meteorological Agency best track data), (c) ENP [1971–2012; reanalysis-derived CPS ET designations by Wood and Ritchie (2014a)], and (d) SWIO [1987–2013; reanalysis-derived ET designations subjectively determined by Griffin and Bosart (2014)]. No attempt is made to account for ET classification practice differences between operational centers or the historical evolution of ET classification practices at these centers.
Conceptual model of the transformation stage of ET in the western North Pacific. Step 1 represents the commencement of the transformation stage, step 2 represents the TC encountering the baroclinic zone, and step 3 represents the TC becoming embedded within the baroclinic zone. Figure reproduced from Klein et al. (2000, their Fig. 5).
Number of TCs (white bars) and number of ET cases (black bars) by month (each per left axis) during 1979–2004 in the WNP, as assessed using Japan Meteorological Agency best track data. The black line indicates the percentage of TCs that undergo ET to the total number of TCs in a given month (per right axis). Figure reproduced from Kitabatake (2011, their Fig. 10a).
Average JRA-55 CPS frequency (e.g., number of times during its life span that a given TC is located at a given location within the CPS, 
Summary of TC and ET events in the SWIO west of 90°E by (a) TC season and (b) month for the period 1989–2013. The full height of the bar represents TC events, while the bottom (blue) portion of the bar represents the number of ET events. In (a), the year on the chart refers to the year the TC season ended. In (b), events that occur in two months are included in the month in which the TC dissipated or underwent ET. Figure reproduced from Griffin and Bosart (2014, their Fig. 1).
Azimuthally averaged 10-m wind speed (m s−1) as a function of radius at 0400 UTC 29 Aug (open circles; before ET), 1000 UTC 30 Aug (closed circles; during ET), and 1000 UTC 31 Aug (open squares; after ET) 1998, as obtained from the 12-km fifth-generation Pennsylvania State University–NCAR Mesoscale Model (Dudhia 1993), simulation of NATL TC Bonnie (1998). Figure reproduced from Evans and Hart (2008, their Fig. 5).
Composite surface wind vectors (arrows, reference vector in the top right of each panel) and surface wind speed (isotachs, m s−1) for (a) the subset of TCs (n = 13) with wind maxima both left and right of track that made landfall in Japan from 1979 to 2004 and (b) all TCs (n = 70) that made landfall in Japan from 1979 to 2004. The y axis is taken in the direction of the storm motion. The cross in the center of each panel indicates the storm center. Figure reproduced from Fujibe and Kitabatake (2007, their Figs. 3d,f).
Vertical cross sections of radar reflectivity (dBZ; shaded), tangential wind (m s−1; gray contours), and in-plane wind vectors composed of radial and vertical velocity (m s−1) for each shear-relative quadrant as synthesized by the Spline Analysis at Mesoscale Utilizing Radar and Aircraft Instrumentation (SAMURAI; Bell et al. 2012) software tool for the T-PARC research mission into Typhoon Sinlaku on 19 Sep 2008. Cross sections are taken from the corner of the domain to the center, 45° from the x and y axes in each quadrant: (a) upshear left, (b) downshear left, (c) upshear right, and (d) downshear right. Figure reproduced from Foerster et al. (2014, their Fig. 12).
Maximum reflectivity between 0 and 15 km (dBZ; shaded), temperature (K; contours) at 1.5 km, and horizontal wind (m s−1; vectors with reference vector at lower right) at 1.5 km as synthesized by SAMURAI for the T-PARC research mission into Typhoon Sinlaku on 20 Sep 2008. The gray line denotes the flight track of the NRL-P3 and the red line denotes the flight track of the USAF-WC130. Filled circles give positions of dropsondes included in the SAMURAI analysis. Stars indicate positions of dropsondes in Fig. 7 in Quinting et al. (2014), while black arrows along the coordinate axes indicate positions of cross sections in Fig. 6 in Quinting et al. (2014). Figure reproduced from Quinting et al. (2014, their Fig. 4).
Advanced Hurricane Weather Research and Forecasting (AHW; Davis et al. 2010) model-forecast 850-hPa potential temperature (K; shaded), vector wind (half barb = 2.5 m s−1; full barb = 5.0 m s−1; pennant = 25.0 m s−1), and wind speed (solid contours at 40, 45, 50, and 55 m s−1) of Sandy (2012) verifying at (a) 1000 and (b) 2000 UTC 29 Oct 2012. The AHW forecast was initialized at 0000 UTC 28 Oct 2012. Figure reproduced from Galarneau et al. (2013, their Figs. 7b,c).
12-h forward trajectories starting at 925 hPa on a northwest–southeast-directed line crossing Japan at (a) 0000 UTC 19 Sep 2008 and (b) 0000 UTC 20 Sep 2008. The colors of the trajectories represent pressure (hPa). Equivalent potential temperature (K) at 925 hPa at trajectory starting time is given in gray shades. The location of Sinlaku’s simulated mean sea level pressure minimum is marked by a blue cross (at trajectory start) and circle (at trajectory end). Figure reproduced from Lentink (2017, their Figs. 5.3 and 5.20b).
The average absolute and along- and cross-track errors of the NCEP Global Ensemble Forecast System in the NATL and WNP basins for the period 2006–08. Error bars illustrate 95% confidence intervals on the mean as determined using bootstrapping. Both TC and ET tracks are included in the analysis. Along-track error is positive when a forecast lies ahead of its verifying position and cross-track error is positive when a cyclone is forecast to the right of its verifying position. Figure reproduced from Buckingham et al. (2010, their Fig. 4). (b) As in (a), but only TC tracks are included in the analysis (ET tracks excluded). Figure reproduced from Buckingham et al. (2010, their Fig. 5).
Percentage of correctly classified cyclone phase forecasts by the linear discriminant analysis scheme of Aberson (2014) for dependent (short dashed; period of record 1980–2010) and independent (medium dashed; period of record 2011) samples. The long-dashed line indicates the percentage of correctly classified official NHC cyclone phase forecasts (period of record 2011). Note that the two 2011 samples are homogeneous. Figure reproduced from Aberson (2014, their Fig. 3).
Selected regional prediction system forecasts for NATL Hurricane Juan initialized at 0000 UTC 28 Sep 2003. Sea level pressure (solid lines; 4-hPa intervals) and winds (barbs; m s−1) are shown for the (left) initial state and (right) 24-h forecasts valid at 0000 UTC 29 Sep 2003. Model fields are indicated for the (a),(b) NCEP ETA Model; (c),(d) regional version of the Environment Canada Global Environmental Multiscale model (GEM-R); (e),(f) GFDL Hurricane Model (GHM); and (g),(h) Mesoscale Compressible Community (MC2) model. Minimum MSLP contours are 984 hPa in (e) and (f). Please see McTaggart-Cowan et al. (2006a) for relevant model details. Figure reproduced from McTaggart-Cowan et al. (2006a, their Fig. 6).
