On stationarizability for nonstationary 2-D random fields using discrete wavelet transforms

Area of Science:

• Stochastic processes and random field theory.
• Signal processing and time-frequency analysis.
• Mathematical physics and statistical mechanics.

Background:

• Focuses on nonstationary two-dimensional (2-D) random fields.
• Investigates fields with wide-sense stationary increments and jumps.
• Examines 2-D fractional Brownian motion (fBm) fields.

Discussion:

• Develops a realizable method for stationarizing nonstationary 2-D random fields.
• Presents correlation functions for discrete wavelet transform (DWT) of 2-D fBm fields.
• Highlights the hyperbolic decay of these correlation functions.

Key Insights:

• A novel stationarization technique is proposed for complex 2-D random fields.
• The discrete wavelet transform reveals specific decay properties in 2-D fBm fields.
• Understanding these properties is crucial for accurate modeling and analysis.

Outlook:

• Potential applications in image processing, geostatistics, and turbulence modeling.
• Further research into other nonstationary field types and transform methods.
• Exploration of the theoretical implications of hyperbolic decay in random field correlations.