Search Results
You are looking at 41 - 50 of 83 items for
- Author or Editor: Michael K. Tippett x
- Refine by Access: All Content x
Abstract
The climate of Saudi Arabia is arid–semiarid with infrequent but sometimes intense rainfall, which can cause flooding. Interannual and intraseasonal precipitation variability in the region is related to ENSO and MJO tropical convection. The predictability of these tropical signals gives some expectation of skillful extended-range rainfall forecasts in the region. Here, the extent to which this predictability is realizable in the Climate Forecast System (CFS), version 2, a state-of-the-art coupled global ocean–atmosphere model, is assessed. While there are deficiencies in the forecast climatology likely related to orography and resolution, as well as lead-dependent biases, CFS represents the climatology of the region reasonably well. Forecasts of the areal average of rainfall over Saudi Arabia show that the CFS captures some features of a spring 2013 heavy rainfall event up to 10 days in advance and a transition from dry to wet conditions up to 20 days in advance. Analysis of a 12-yr (1999–2010) reforecast dataset shows that the CFS can skillfully predict the rainfall amount, the number of days exceeding a threshold, and the probability of heavy rainfall occurrence for forecast windows ranging from 1 to 30 days. While the probability forecasts show good discrimination, they are overconfident. Logistic regression based on the ensemble mean value improves forecast skill and reliability. Forecast probabilities have a clear relation with the MJO phase in the wet season, providing a physical basis for the observed forecast skill.
Abstract
The climate of Saudi Arabia is arid–semiarid with infrequent but sometimes intense rainfall, which can cause flooding. Interannual and intraseasonal precipitation variability in the region is related to ENSO and MJO tropical convection. The predictability of these tropical signals gives some expectation of skillful extended-range rainfall forecasts in the region. Here, the extent to which this predictability is realizable in the Climate Forecast System (CFS), version 2, a state-of-the-art coupled global ocean–atmosphere model, is assessed. While there are deficiencies in the forecast climatology likely related to orography and resolution, as well as lead-dependent biases, CFS represents the climatology of the region reasonably well. Forecasts of the areal average of rainfall over Saudi Arabia show that the CFS captures some features of a spring 2013 heavy rainfall event up to 10 days in advance and a transition from dry to wet conditions up to 20 days in advance. Analysis of a 12-yr (1999–2010) reforecast dataset shows that the CFS can skillfully predict the rainfall amount, the number of days exceeding a threshold, and the probability of heavy rainfall occurrence for forecast windows ranging from 1 to 30 days. While the probability forecasts show good discrimination, they are overconfident. Logistic regression based on the ensemble mean value improves forecast skill and reliability. Forecast probabilities have a clear relation with the MJO phase in the wet season, providing a physical basis for the observed forecast skill.
ABSTRACT
Rapid intensification (RI) is an outstanding source of error in tropical cyclone (TC) intensity predictions. RI is generally defined as a 24-h increase in TC maximum sustained surface wind speed greater than some threshold, typically 25, 30, or 35 kt (1 kt ≈ 0.51 m s−1). Here, a long short-term memory (LSTM) model for probabilistic RI predictions is developed and evaluated. The variables (features) of the model include storm characteristics (e.g., storm intensity) and environmental variables (e.g., vertical shear) over the previous 48 h. A basin-aware RI prediction model is trained (1981–2009), validated (2010–13), and tested (2014–17) on global data. Models are trained on overlapping 48-h data, which allows multiple training examples for each storm. A challenge is that the data are highly unbalanced in the sense that there are many more non-RI cases than RI cases. To cope with this data imbalance, the synthetic minority-oversampling technique (SMOTE) is used to balance the training data by generating artificial RI cases. Model ensembling is also applied to improve prediction skill further. The model’s Brier skill scores in the Atlantic and eastern North Pacific are higher than those of operational predictions for RI thresholds of 25 and 30 kt and comparable for 35 kt on the independent test data. Composites of the features associated with RI and non-RI situations provide physical insights for how the model discriminates between RI and non-RI cases. Prediction case studies are presented for some recent storms.
ABSTRACT
Rapid intensification (RI) is an outstanding source of error in tropical cyclone (TC) intensity predictions. RI is generally defined as a 24-h increase in TC maximum sustained surface wind speed greater than some threshold, typically 25, 30, or 35 kt (1 kt ≈ 0.51 m s−1). Here, a long short-term memory (LSTM) model for probabilistic RI predictions is developed and evaluated. The variables (features) of the model include storm characteristics (e.g., storm intensity) and environmental variables (e.g., vertical shear) over the previous 48 h. A basin-aware RI prediction model is trained (1981–2009), validated (2010–13), and tested (2014–17) on global data. Models are trained on overlapping 48-h data, which allows multiple training examples for each storm. A challenge is that the data are highly unbalanced in the sense that there are many more non-RI cases than RI cases. To cope with this data imbalance, the synthetic minority-oversampling technique (SMOTE) is used to balance the training data by generating artificial RI cases. Model ensembling is also applied to improve prediction skill further. The model’s Brier skill scores in the Atlantic and eastern North Pacific are higher than those of operational predictions for RI thresholds of 25 and 30 kt and comparable for 35 kt on the independent test data. Composites of the features associated with RI and non-RI situations provide physical insights for how the model discriminates between RI and non-RI cases. Prediction case studies are presented for some recent storms.
Abstract
The relation between the El Niño–Southern Oscillation (ENSO) and California precipitation has been studied extensively and plays a prominent role in seasonal forecasting. However, a wide range of precipitation outcomes on seasonal time scales are possible, even during extreme ENSO states. Here, we investigate prediction skill and its origins on subseasonal time scales. Model predictions of California precipitation are examined using Subseasonal Experiment (SubX) reforecasts for the period 1999–2016, focusing on those from the Flow-Following Icosahedral Model (FIM). Two potential sources of subseasonal predictability are examined: the tropical Pacific Ocean and upper-level zonal winds near California. In both observations and forecasts, the Niño-3.4 index exhibits a weak and insignificant relationship with daily to monthly averages of California precipitation. Likewise, model tropical sea surface temperature and outgoing longwave radiation show only minimal relations with California precipitation forecasts, providing no evidence that flavors of El Niño or tropical modes substantially contribute to the success or failure of subseasonal forecasts. On the other hand, an index for upper-level zonal winds is strongly correlated with precipitation in observations and forecasts, across averaging windows and lead times. The wind index is related to ENSO, but the correlation between the wind index and precipitation remains even after accounting for ENSO phase. Intriguingly, the Niño-3.4 index and California precipitation show a slight but robust negative statistical relation after accounting for the wind index.
Abstract
The relation between the El Niño–Southern Oscillation (ENSO) and California precipitation has been studied extensively and plays a prominent role in seasonal forecasting. However, a wide range of precipitation outcomes on seasonal time scales are possible, even during extreme ENSO states. Here, we investigate prediction skill and its origins on subseasonal time scales. Model predictions of California precipitation are examined using Subseasonal Experiment (SubX) reforecasts for the period 1999–2016, focusing on those from the Flow-Following Icosahedral Model (FIM). Two potential sources of subseasonal predictability are examined: the tropical Pacific Ocean and upper-level zonal winds near California. In both observations and forecasts, the Niño-3.4 index exhibits a weak and insignificant relationship with daily to monthly averages of California precipitation. Likewise, model tropical sea surface temperature and outgoing longwave radiation show only minimal relations with California precipitation forecasts, providing no evidence that flavors of El Niño or tropical modes substantially contribute to the success or failure of subseasonal forecasts. On the other hand, an index for upper-level zonal winds is strongly correlated with precipitation in observations and forecasts, across averaging windows and lead times. The wind index is related to ENSO, but the correlation between the wind index and precipitation remains even after accounting for ENSO phase. Intriguingly, the Niño-3.4 index and California precipitation show a slight but robust negative statistical relation after accounting for the wind index.
Abstract
Two questions are addressed in this paper: whether ENSO can be adequately characterized by simple, seasonally invariant indices and whether the time series of a single component—SST or OLR—provides a sufficiently complete representation of ENSO for the purpose of quantifying U.S. climate impacts. Here, ENSO is defined as the leading mode of seasonally varying canonical correlation analysis (CCA) between anomalies of tropical Pacific SST and outgoing longwave radiation (OLR). The CCA reveals that the strongest regions of coupling are mostly invariant as a function of season and correspond to an OLR region located in the central Pacific Ocean (CP-OLR) and an SST region in the eastern Pacific that coincides with the Niño-3 region. In a linear context, the authors explore whether the use of a combined index of these SST and OLR regions explains additional variance of North American temperature and precipitation anomalies beyond that described by using a single index alone. Certain seasons and regions benefit from the use of a combined index. In particular, a combined index describes more variability in winter/spring precipitation and summer temperature.
Abstract
Two questions are addressed in this paper: whether ENSO can be adequately characterized by simple, seasonally invariant indices and whether the time series of a single component—SST or OLR—provides a sufficiently complete representation of ENSO for the purpose of quantifying U.S. climate impacts. Here, ENSO is defined as the leading mode of seasonally varying canonical correlation analysis (CCA) between anomalies of tropical Pacific SST and outgoing longwave radiation (OLR). The CCA reveals that the strongest regions of coupling are mostly invariant as a function of season and correspond to an OLR region located in the central Pacific Ocean (CP-OLR) and an SST region in the eastern Pacific that coincides with the Niño-3 region. In a linear context, the authors explore whether the use of a combined index of these SST and OLR regions explains additional variance of North American temperature and precipitation anomalies beyond that described by using a single index alone. Certain seasons and regions benefit from the use of a combined index. In particular, a combined index describes more variability in winter/spring precipitation and summer temperature.
Abstract
Here, we examine the relation between U.S. tornado activity and a new year-round classification of North American weather regimes. The regime classification is based on 500-hPa geopotential height anomalies and classifies each day as Pacific Trough, Pacific Ridge, Alaskan Ridge, Greenland High, or No regime. During the period 1979–2022, we find statistically significant relations between average tornado report numbers and weather regimes in all months except June–August. Tornado activity is enhanced on Pacific Ridge days during late winter and spring, reduced on Pacific Trough days in spring, and reduced on Alaskan Ridge and Greenland High days during fall and early winter. During active regimes, the probability of many tornadoes occurring also increases, and there is greater variability in the number of tornadoes reported each day. A reanalysis-based tornado index reproduces the regional features of the modulation of tornado activity by the weather regimes and attributes them to changes in storm relative helicity, convective available potential energy, and convective precipitation. The phase of El Niño–Southern Oscillation (ENSO) also plays a role. In winter and spring, Pacific Ridge days occur more often and average more reports per day during cool ENSO conditions. During warm ENSO conditions, Pacific Trough days occur more often and are associated with widespread reduced tornado activity.
Significance Statement
Daily weather patterns over North America can be classified into five categories. The purpose of this study was to examine whether the number of U.S. tornado reports on a given day depends on the weather category of that day. We found robust relations between the average number of tornado reports and the weather pattern category in all months except June–August, with some weather patterns associated with increased tornado numbers and others with decreased tornado numbers. The El Niño–Southern Oscillation (ENSO) phenomenon plays a role, with weather patterns that are favorable for tornadoes being more frequent and having more tornadoes per day during cool ENSO conditions.
Abstract
Here, we examine the relation between U.S. tornado activity and a new year-round classification of North American weather regimes. The regime classification is based on 500-hPa geopotential height anomalies and classifies each day as Pacific Trough, Pacific Ridge, Alaskan Ridge, Greenland High, or No regime. During the period 1979–2022, we find statistically significant relations between average tornado report numbers and weather regimes in all months except June–August. Tornado activity is enhanced on Pacific Ridge days during late winter and spring, reduced on Pacific Trough days in spring, and reduced on Alaskan Ridge and Greenland High days during fall and early winter. During active regimes, the probability of many tornadoes occurring also increases, and there is greater variability in the number of tornadoes reported each day. A reanalysis-based tornado index reproduces the regional features of the modulation of tornado activity by the weather regimes and attributes them to changes in storm relative helicity, convective available potential energy, and convective precipitation. The phase of El Niño–Southern Oscillation (ENSO) also plays a role. In winter and spring, Pacific Ridge days occur more often and average more reports per day during cool ENSO conditions. During warm ENSO conditions, Pacific Trough days occur more often and are associated with widespread reduced tornado activity.
Significance Statement
Daily weather patterns over North America can be classified into five categories. The purpose of this study was to examine whether the number of U.S. tornado reports on a given day depends on the weather category of that day. We found robust relations between the average number of tornado reports and the weather pattern category in all months except June–August, with some weather patterns associated with increased tornado numbers and others with decreased tornado numbers. The El Niño–Southern Oscillation (ENSO) phenomenon plays a role, with weather patterns that are favorable for tornadoes being more frequent and having more tornadoes per day during cool ENSO conditions.
Abstract
Ensemble data assimilation methods assimilate observations using state-space estimation methods and low-rank representations of forecast and analysis error covariances. A key element of such methods is the transformation of the forecast ensemble into an analysis ensemble with appropriate statistics. This transformation may be performed stochastically by treating observations as random variables, or deterministically by requiring that the updated analysis perturbations satisfy the Kalman filter analysis error covariance equation. Deterministic analysis ensemble updates are implementations of Kalman square root filters. The nonuniqueness of the deterministic transformation used in square root Kalman filters provides a framework to compare three recently proposed ensemble data assimilation methods.
Abstract
Ensemble data assimilation methods assimilate observations using state-space estimation methods and low-rank representations of forecast and analysis error covariances. A key element of such methods is the transformation of the forecast ensemble into an analysis ensemble with appropriate statistics. This transformation may be performed stochastically by treating observations as random variables, or deterministically by requiring that the updated analysis perturbations satisfy the Kalman filter analysis error covariance equation. Deterministic analysis ensemble updates are implementations of Kalman square root filters. The nonuniqueness of the deterministic transformation used in square root Kalman filters provides a framework to compare three recently proposed ensemble data assimilation methods.
Abstract
Here we present a machine learning–based wind reconstruction model. The model reconstructs hurricane surface winds with XGBoost, which is a decision-tree-based ensemble predictive algorithm. The model treats the symmetric and asymmetric wind fields separately. The symmetric wind field is approximated by a parametric wind profile model and two Bessel function series. The asymmetric field, accounting for asymmetries induced by the storm and its ambient environment, is represented using a small number of Laplacian eigenfunctions. The coefficients associated with Bessel functions and eigenfunctions are predicted by XGBoost based on storm and environmental features taken from NHC best-track and ERA-Interim data, respectively. We use HWIND for the observed wind fields. Three parametric wind profile models are tested in the symmetric wind model. The wind reconstruction model’s performance is insensitive to the choice of the profile model because the Bessel function series correct biases of the parametric profiles. The mean square error of the reconstructed surface winds is smaller than the climatological variance, indicating skillful reconstruction. Storm center location, eyewall size, and translation speed play important roles in controlling the magnitude of the leading asymmetries, while the phase of the asymmetries is mainly affected by storm translation direction. Vertical wind shear impacts the asymmetry phase to a lesser degree. Intended applications of this model include assessing hurricane risk using synthetic storm event sets generated by statistical–dynamical downscaling hurricane models.
Abstract
Here we present a machine learning–based wind reconstruction model. The model reconstructs hurricane surface winds with XGBoost, which is a decision-tree-based ensemble predictive algorithm. The model treats the symmetric and asymmetric wind fields separately. The symmetric wind field is approximated by a parametric wind profile model and two Bessel function series. The asymmetric field, accounting for asymmetries induced by the storm and its ambient environment, is represented using a small number of Laplacian eigenfunctions. The coefficients associated with Bessel functions and eigenfunctions are predicted by XGBoost based on storm and environmental features taken from NHC best-track and ERA-Interim data, respectively. We use HWIND for the observed wind fields. Three parametric wind profile models are tested in the symmetric wind model. The wind reconstruction model’s performance is insensitive to the choice of the profile model because the Bessel function series correct biases of the parametric profiles. The mean square error of the reconstructed surface winds is smaller than the climatological variance, indicating skillful reconstruction. Storm center location, eyewall size, and translation speed play important roles in controlling the magnitude of the leading asymmetries, while the phase of the asymmetries is mainly affected by storm translation direction. Vertical wind shear impacts the asymmetry phase to a lesser degree. Intended applications of this model include assessing hurricane risk using synthetic storm event sets generated by statistical–dynamical downscaling hurricane models.
Abstract
This paper introduces a logistic regression model for the extratropical transition (ET) of tropical cyclones in the North Atlantic and the western North Pacific, using elastic net regularization to select predictors and estimate coefficients. Predictors are chosen from the 1979–2017 best track and reanalysis datasets, and verification is done against the tropical/extratropical labels in the best track data. In an independent test set, the model skillfully predicts ET at lead times up to 2 days, with latitude and sea surface temperature as its most important predictors. At a lead time of 24 h, it predicts ET with a Matthews correlation coefficient of 0.4 in the North Atlantic, and 0.6 in the western North Pacific. It identifies 80% of storms undergoing ET in the North Atlantic and 92% of those in the western North Pacific. In total, 90% of transition time errors are less than 24 h. Select examples of the model’s performance on individual storms illustrate its strengths and weaknesses. Two versions of the model are presented: an “operational model” that may provide baseline guidance for operational forecasts and a “hazard model” that can be integrated into statistical TC risk models. As instantaneous diagnostics for tropical/extratropical status, both models’ zero lead time predictions perform about as well as the widely used cyclone phase space (CPS) in the western North Pacific and better than the CPS in the North Atlantic, and predict the timings of the transitions better than CPS in both basins.
Abstract
This paper introduces a logistic regression model for the extratropical transition (ET) of tropical cyclones in the North Atlantic and the western North Pacific, using elastic net regularization to select predictors and estimate coefficients. Predictors are chosen from the 1979–2017 best track and reanalysis datasets, and verification is done against the tropical/extratropical labels in the best track data. In an independent test set, the model skillfully predicts ET at lead times up to 2 days, with latitude and sea surface temperature as its most important predictors. At a lead time of 24 h, it predicts ET with a Matthews correlation coefficient of 0.4 in the North Atlantic, and 0.6 in the western North Pacific. It identifies 80% of storms undergoing ET in the North Atlantic and 92% of those in the western North Pacific. In total, 90% of transition time errors are less than 24 h. Select examples of the model’s performance on individual storms illustrate its strengths and weaknesses. Two versions of the model are presented: an “operational model” that may provide baseline guidance for operational forecasts and a “hazard model” that can be integrated into statistical TC risk models. As instantaneous diagnostics for tropical/extratropical status, both models’ zero lead time predictions perform about as well as the widely used cyclone phase space (CPS) in the western North Pacific and better than the CPS in the North Atlantic, and predict the timings of the transitions better than CPS in both basins.