A Fractal Analysis of Near-Field Atmospheric Concentration Data
Abstract
A previous study showed that regional and large scale atmospheric motions have Hurst exponents on the order of .4 to .5. The Hurst exponent, H, equals 2-D, where D is the fractal dimension, for a time series. These numbers characterize the degree of irregularity of the time-series. Hurst exponents were calculated for a large number of concentration-time series from several sources, whose sampling times ranged from 6 seconds to 1 hour. Hurst exponents for these data ranged from about .2 to .4. Such values indicate that the small-scale, near-field turbulent eddies differ qualitively from the larger-scale motions, for which H approaches the value .5, the value for Brownian motion. Since models in effect assume the value H = .5 in turbulence parameterizations, these, results suggest that smaller, realistic H-values, appropriate to planetary boundary layer turbulence, should be used to parameterize turbulence.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 25, 1993
- Accession Number
- ADA275460
Entities
People
- Franklin A. Gifford