The Hurst exponent is used as a measure of long-term memory of time series. The Rescaled Range Analysis (R/S method) is best known for estimation of Hurst exponent. I followed all its steps, compared my code results with those available at mathworks.com, but didn't find agreement among the results of these codes. The links for these codes are:
https://in.mathworks.com/matlabcentral/fileexchange/9842-hurst-exponent?s_tid=prof_contriblnk
https://in.mathworks.com/matlabcentral/fileexchange/57388-hurst-exponent-and-linear-fit
https://in.mathworks.com/matlabcentral/fileexchange/25414-rescaled-range-analysis
Can anyone help me out with this? I feel difficulty regarding estimation of Hurst coefficient.
Thanks in advance!
Here is a reference to a paper in which we used R/S to calculate the Hurst exponent for stratospheric variables: A F Tuck & S J Hovde, Geophysical Research Letters, 26, 1271-1274[1999]. Since 1999, the use of R/S has been updated by the generalised scale invariance (GSI) methods originated by Schertzer & Lovejoy. Matlab works on single precision, and we found that our results were very sensitive until we switched to double precision (ie 64 rather than 32 bits/word). GSI may be worth trying, it has given us greatly improved analysis of fractal behaviour in natural geophysical time series.
Thank you very much Sir!!
I will surely try this and get back to you in case I face any query!
No. The Hurst exponent is a part of GSI, which has moved on from rescaled range. Detailed explanations can be found in the following references.
S Lovejoy & D Schertzer, 2013, THE WEATHER AND CLIMATE: Emergent Laws and Multifractal Cascades, CUP, ISBN 9781107018983
A F Tuck, 2008, ATMOSPHERIC TURBULENCE: A Molecular Dynamics Perspective,
OUP, ISBN 9780199236534
A F Tuck, 2010, Quart. J. R. Meteorol. Soc., 136, 1125-1144. From molecules to meteorology via turbulent scale invariance. Doi: 10.1002/qj.644
- Badges
- Science topic
- Similar topics
- Geoscience
- Hydrology