Reducing the Dynamic State Index to its main information using Principal Component Analysis
Authors:
Annette Müller,
Nikolai Spaeth,
Johanna Bollow,
Iris Eder,
Alina Fischer,
Jana Frenzel,
Annika Görnt,
Kunyan Hao,
Johanne Ilchmann,
Isabel Kasner,
Corinna Langwald,
Steffen Laube,
Flavio Maggioni,
Beatrice Neiszt-Pllana,
Daniel A. Redant,
Lisa Rogge,
Sus Sama,
David Scheer,
Inga Spreen,
Patrick W. Varchmin,
Vanessa R. Zabel,
Wei Zhang,
Anika Zibell
Abstract:
The Dynamic State Index is a scalar quantity designed to identify atmospheric developments such as fronts, hurricanes or specific weather pattern. The DSI is defined as Jacobian-determinant of three constitutive quantities that characterize three-dimensional fluid flows: the Bernoulli stream function, the potential vorticity (PV) and the potential temperature. Here, we tackle the questions (i) if…
▽ More
The Dynamic State Index is a scalar quantity designed to identify atmospheric developments such as fronts, hurricanes or specific weather pattern. The DSI is defined as Jacobian-determinant of three constitutive quantities that characterize three-dimensional fluid flows: the Bernoulli stream function, the potential vorticity (PV) and the potential temperature. Here, we tackle the questions (i) if the mathematical formulation of the DSI can be reduced, while keeping the main information, and (ii) does the reduction of the DSI depend on the spatial scale? Applying principle component analysis we find that three of six DSI terms that sum up to the Jacobi-determinant are sufficient for future DSI calculations.
△ Less
Submitted 6 March, 2022;
originally announced March 2022.
