Forecasting Zonda Wind Occurrence with Vertical Sounding Data

Otero, Federico; Araneo, Diego Christian

doi:10.1007/s00376-021-1007-0

Cited by 1 publication

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Otero and Araneo (2022) presented a probabilistic model for the Zonda occurrence constructed from the logistic correlation between an observed occurrence index and the PCA component loadings of a predictor variable (vertical temperature, dew point, wind and stability profiles, from vertical sounding data). Following these ideas, in this work we will use the ERA5 reanalysis fields of different meteorological variables as predictors.…”

Section: Methodsmentioning

confidence: 99%

“…The methodology for classifying Zonda events follows the approach used in Otero and Araneo (2022), which resulted in a high‐performance model using machine learning algorithms. The predictor variables employed in this study are temperature ( T ), dew point temperature (Td), surface pressure ( P ) and 10‐m wind speed ( V ).…”

Section: Methodsmentioning

confidence: 99%

“…This model was capable of discriminating between Zonda and non‐Zonda events with an effectiveness close to 91%, a probability of detection above 97%, a probability of false detection lower than 7% and a missing ratio below 1%. In the present work, the methodology of Otero and Araneo (2022) is followed. In contrast to this earlier work, which only took into account the vertical sounding on both sides of the Andes, in this instance, spatial reanalysis fields are used.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Synoptic fingerprints of Zonda wind from a statistical prediction model

Otero,

Araneo

2023

Intl Journal of Climatology

Self Cite

View full text Add to dashboard Cite

Zonda wind is a typical downslope windstorm over the eastern slopes of the Central Andes in Argentina, which produces extremely warm and dry conditions and has substantial socioeconomic impacts. In this study, we propose a new statistical model for Zonda prediction based on the “synoptic fingerprints” of atmospheric diagnostic variables from ERA5. The model combines principal component analysis (PCA) and logistic regression to establish a relationship between the observed occurrence and the PCA loading component of a predictor variable. This approach enables us to determine the probability of Zonda occurrence at selected stations and identify the synoptic structure features (fingerprints) associated with Zonda events. The obtained fields successfully discriminate between Zonda and non‐Zonda events, suggesting that the available information in the reanalysis data is sufficient for predicting the presence of Zonda. The synoptic fingerprints generated by the model reveal a cross‐barrier pressure gradient resulting from a negative geopotential height anomaly at low levels. The cross‐barrier flow remains unimpeded by the Andes, leading to forced vertical motions on the windward side, accompanied by cooling and precipitation. On the lee side, sinking motions, warming and drying are observed, further facilitated by favourable mid‐ and upper‐level conditions that establish the Zonda wind. The model performs comparably to previous research, with the best results achieved using low‐level variables as predictors. Key performance measures, including the area under the receiver operating curve (ROC) (AUC) of 0.9468 and a Brier score lower than 0.1, demonstrate the model's effectiveness. Using a 0.5 threshold, the accuracy, F1 score and correct alarm ratio (CAR) all exceed 88%, with a probability of detection (POD) higher than 90%. Studies on this type of downslope windstorm are scarce in South America, making this work a significant contribution to understanding synoptic‐scale atmospheric structures associated with Zonda occurrences.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation