Abstract

A 42-yr study of eastern North Pacific tropical cyclones (TCs) undergoing extratropical transition (ET) is presented using the Japanese 55-yr Reanalysis dataset. By using cyclone phase space (CPS) to differentiate those TCs that undergo ET from those that do not, it is found that only 9% of eastern North Pacific TCs that developed from 1971 to 2012 complete ET, compared with 40% in the North Atlantic.

Using a combination of CPS, empirical orthogonal function (EOF) analysis, and composite analysis, it is found that the evolution of ET in this basin differs from that observed in the North Atlantic and western North Pacific, possibly as a result of the rapidly decreasing sea surface temperatures north of the main genesis region. The presence of a strong, deep subtropical ridge extending westward from North America into the eastern North Pacific is a major factor inhibiting ET in this basin. Similar to other basins, eastern North Pacific ET generally occurs in conjunction with an approaching midlatitude trough, which helps to weaken the ridge and allow northward passage of the TC. The frequency of ET appears to increase during developing El Niño events but is not significantly affected by the Pacific decadal oscillation.

Abstract

A dataset of 167 eastern North Pacific tropical cyclones (TCs) is investigated for potential impacts in the southwestern United States over the period 1989–2009 and evaluated in the context of a 30-yr climatology. The statistically significant patterns from empirical orthogonal function (EOF) analysis demonstrate the prevalence of a midlatitude trough pattern when TC-related rainfall occurs in the southwestern United States. Conversely, the presence of a strong subtropical ridge tends to prevent such events from occurring and limits TC-related rainfall to Mexico. These statistically significant patterns correspond well with previous work. The El Niño–Southern Oscillation phenomenon is shown to have some effect on eastern North Pacific TC impacts on the southwestern United States, as shifts in the general circulation can subsequently influence which regions receive rainfall from TCs or their remnants. The Pacific decadal oscillation may have a greater influence during the period of study as evidenced by EOF analysis of sea surface temperature anomalies.

Abstract

A case study of eastern North Pacific Tropical Storm Ignacio (1997), which brought rainfall to the southwestern United States as a tropical cyclone and to the northwestern United States as an extratropical cyclone, is presented. This tropical cyclone formed from a region of disturbed weather, rather than a tropical wave, outside the typical eastern North Pacific genesis region and intensified into a tropical storm coincident with the passage of an upper-tropospheric trough. Moisture transported from Ignacio along an outflow jet associated with the trough resulted in precipitation in Mexico and the southwestern United States. As Ignacio moved north and away from the trough, this tropical cyclone weakened and eventually underwent extratropical transition over the open ocean, in contrast to climatological eastern North Pacific tropical cyclone behavior. Ignacio then strengthened as an extratropical cyclone due to favorable baroclinic conditions and the passage of another upper-tropospheric trough before making landfall on the northern coast of California, bringing rain to the northwestern United States. Ignacio’s remnant moisture eventually merged into a slow-moving midlatitude low pressure system that developed after interacting with the extratropical remnant of Hurricane Guillermo.

Abstract

Tropical cyclone (TC) track forecasts have improved in recent decades while intensity forecasts, particularly predictions of rapid intensification (RI), continue to show low skill. Many statistical methods have shown promise in predicting RI using environmental fields, although these methods rely heavily upon supervised learning techniques such as classification. Advances in unsupervised learning techniques, particularly those that integrate nonlinearity into the class separation problem, can improve discrimination ability for difficult tasks such as RI prediction. This study quantifies separability between RI and non-RI environments for 2004–16 Atlantic Ocean TCs using an unsupervised learning method that blends principal component analysis with k-means cluster analysis. Input fields consisted of TC-centered 1° Global Forecast System analysis (GFSA) grids (170 different variables and isobaric levels) for 3605 TC samples and five domain sizes. Results are directly compared with separability offered by operational RI forecast predictors for eight RI definitions. The unsupervised learning procedure produced improved separability over operational predictors for all eight RI definitions, five of which showed statistically significant improvement. Composites from these best-separating GFSA fields highlighted the importance of mid- and upper-level relative humidity in identifying the onset of short-term RI, whereas long-term, higher-magnitude RI was generally associated with weaker absolute vorticity. Other useful predictors included optimal thermodynamic RI ingredients along the mean trajectory of the TC. The results suggest that the orientation of a more favorable thermodynamic environment relative to the TC and midlevel vorticity magnitudes could be useful predictors for RI.

Abstract

Tropical cyclones (TCs) rank among the most costly natural disasters in the United States, and accurate forecasts of track and intensity are critical for emergency response. Intensity guidance has improved steadily but slowly, as processes that drive intensity change are not fully understood. Because most TCs develop far from land-based observing networks, geostationary satellite imagery is critical to monitor these storms. However, these complex data can be challenging to analyze in real time, and off-the-shelf machine-learning algorithms have limited applicability on this front because of their “black box” structure. This study presents analytic tools that quantify convective structure patterns in infrared satellite imagery for overocean TCs, yielding lower-dimensional but rich representations that support analysis and visualization of how these patterns evolve during rapid intensity change. The proposed feature suite targets the global organization, radial structure, and bulk morphology (ORB) of TCs. By combining ORB and empirical orthogonal functions, we arrive at an interpretable and rich representation of convective structure patterns that serve as inputs to machine-learning methods. This study uses the logistic lasso, a penalized generalized linear model, to relate predictors to rapid intensity change. Using ORB alone, binary classifiers identifying the presence (vs absence) of such intensity-change events can achieve accuracy comparable to classifiers using environmental predictors alone, with a combined predictor set improving classification accuracy in some settings. More complex nonlinear machine-learning methods did not perform better than the linear logistic lasso model for current data.

Abstract

Forty-three eastern North Pacific tropical cyclone remnants with varying impact on the southwestern United States during the period 1992–2005 are investigated. Of these, 35 remnants (81%) brought precipitation to some part of the southwestern United States and the remaining 8 remnants (19%) had precipitation that was almost entirely restricted to Mexico, although cloud cover did advect over the southwestern United States in some of these cases. Although the tropical cyclone–strength winds rapidly diminish upon making landfall, these systems still carry a large quantity of tropical moisture and, upon interaction with mountainous topography, are found to drop up to 30% of the local annual precipitation.

Based on common rainfall patterns and large-scale circulation features, the tropical cyclones are grouped into five categories. These include a northern recurving pattern that is more likely to bring rainfall to the southwestern United States; a southern recurving pattern that brings rainfall across northern Mexico and the Gulf Coast region; a largely north and/or northwestward movement pattern that brings rainfall to the west coast of the United States; a group that is blocked from the southwest by a ridge, which limits rainfall to Mexico; and a small group of cases that are not clearly any of the previous four types. Composites of the first four groups are shown and forecasting strategies for each are described.

Abstract

The deviation-angle variance technique (DAV-T), which was introduced in the North Atlantic basin for tropical cyclone (TC) intensity estimation, is adapted for use in the North Pacific Ocean using the “best-track center” application of the DAV. The adaptations include changes in preprocessing for different data sources [Geostationary Operational Environmental Satellite-East (GOES-E) in the Atlantic, stitched GOES-E–Geostationary Operational Environmental Satellite-West (GOES-W) in the eastern North Pacific, and the Multifunctional Transport Satellite (MTSAT) in the western North Pacific], and retraining the algorithm parameters for different basins. Over the 2007–11 period, DAV-T intensity estimation in the western North Pacific results in a root-mean-square intensity error (RMSE, as measured by the maximum sustained surface winds) of 14.3 kt (1 kt ≈ 0.51 m s⁻¹) when compared to the Joint Typhoon Warning Center best track, utilizing all TCs to train and test the algorithm. The RMSE obtained when testing on an individual year and training with the remaining set lies between 12.9 and 15.1 kt. In the eastern North Pacific the DAV-T produces an RMSE of 13.4 kt utilizing all TCs in 2005–11 when compared with the National Hurricane Center best track. The RMSE for individual years lies between 9.4 and 16.9 kt. The complex environment in the western North Pacific led to an extension to the DAV-T that includes two different radii of computation, producing a parametric surface that relates TC axisymmetry to intensity. The overall RMSE is reduced by an average of 1.3 kt in the western North Pacific and 0.8 kt in the eastern North Pacific. These results for the North Pacific are comparable with previously reported results using the DAV for the North Atlantic basin.

Abstract

Because geostationary satellite (Geo) imagery provides a high temporal resolution window into tropical cyclone (TC) behavior, we investigate the viability of its application to short-term probabilistic forecasts of TC convective structure to subsequently predict TC intensity. Here, we present a prototype model which is trained solely on two inputs: Geo infrared imagery leading up to the synoptic time of interest and intensity estimates up to 6 hours prior to that time. To estimate future TC structure, we compute cloud-top temperature radial profiles from infrared imagery and then simulate the evolution of an ensemble of those profiles over the subsequent 12 hours by applying a Deep Autoregressive Generative Model (PixelSNAIL). To forecast TC intensities at hours 6 and 12, we input operational intensity estimates up to the current time (0 h) and simulated future radial profiles up to +12 h into a “nowcasting” convolutional neural network. We limit our inputs to demonstrate the viability of our approach and to enable quantification of value added by the observed and simulated future radial profiles beyond operational intensity estimates alone. Our prototype model achieves a marginally higher error than the National Hurricane Center’s official forecasts despite excluding environmental factors, such as vertical wind shear and sea surface temperature. We also demonstrate that it is possible to reasonably predict short-term evolution of TC convective structure via radial profiles from Geo infrared imagery, resulting in interpretable structural forecasts that may be valuable for TC operational guidance.

Abstract

The deviation angle variance technique (DAV-T) for genesis detection is applied in the western and eastern North Pacific basins. The DAV-T quantifies the axisymmetric organization of cloud clusters using infrared brightness temperature. Since axisymmetry is typically correlated with intensity, the technique can be used to identify relatively high levels of organization at early stages of storm life cycles associated with tropical cyclogenesis. In addition, the technique can be used to automatically track cloud clusters that exhibit signs of organization. In the western North Pacific, automated tracking results for the 2009–11 typhoon seasons show that for a false alarm rate of 25.6%, 96.8% of developing tropical cyclones are detected with a median time of 18.5 h before the cluster reaches an intensity of 30 knots (kt; 1 kt = 0.51 m s⁻¹) in the Joint Typhoon Warning Center best track at a DAV threshold of 1750°². In the eastern North Pacific, for a false alarm rate of 38.0%, the system detects 92.9% of developing tropical cyclones with a median time of 1.25 h before the cluster reaches an intensity of 30 kt in the National Hurricane Center best track during the 2009–11 hurricane seasons at a DAV threshold of 1650°². A significant decrease in tracked nondeveloping clusters occurs when a second organization threshold is introduced, particularly in the western North Pacific.

Abstract

The active 2020 Atlantic hurricane season produced 30 named storms, 14 hurricanes, and 7 major hurricanes (category 3+ on the Saffir–Simpson hurricane wind scale). Though the season was active overall, the final two months (October–November) raised 2020 into the upper echelon of Atlantic hurricane activity for integrated metrics such as accumulated cyclone energy (ACE). This study focuses on October–November 2020, when 7 named storms, 6 hurricanes, and 5 major hurricanes formed and produced ACE of 74 × 10⁴ kt² (1 kt ≈ 0.51 m s⁻¹). Since 1950, October–November 2020 ranks tied for third for named storms, first for hurricanes and major hurricanes, and second for ACE. Six named storms also underwent rapid intensification (≥30 kt intensification in ≤24 h) in October–November 2020—the most on record. This manuscript includes a climatological analysis of October–November tropical cyclones (TCs) and their primary formation regions. In 2020, anomalously low wind shear in the western Caribbean and Gulf of Mexico, likely driven by a moderate-intensity La Niña event and anomalously high sea surface temperatures (SSTs) in the Caribbean, provided dynamic and thermodynamic conditions that were much more conducive than normal for late-season TC formation and rapid intensification. This study also highlights October–November 2020 landfalls, including Hurricanes Delta and Zeta in Louisiana and in Mexico and Hurricanes Eta and Iota in Nicaragua. The active late season in the Caribbean would have been anticipated by a statistical model using the July–September-averaged ENSO longitude index and Atlantic warm pool SSTs as predictors.