Abstract
Time-scale analysis aims at examining the behavior of a stochastic process relative to different observation scales. The need for such an analysis arises often in hydrology where description, modeling and prediction of processes is sought over a large range of spatial and temporal scales. In this paper we present a time-scale analysis framework akin to generalized random processes and based on orthogonal wavelet transforms. In particular, we demonstrate the advantages of using orthogonal wavelets as the kernels in integral processes to obtain optimal averaging and generalized fluctuations (removal of polynomial trends) of stochastic processes over different scales. This paper focuses on one-dimensional processes.
| Original language | English (US) |
|---|---|
| Title of host publication | Finite Elements in Water Resources, Proceedings of the International Conference |
| Publisher | Publ by Computational Mechanics Publ |
| Pages | 99-106 |
| Number of pages | 8 |
| Volume | 2 |
| State | Published - Jan 1 1992 |
| Event | Proceedings of the 9th International Conference on Computational Methods in Water Resources - Denver, CO, USA Duration: Jun 1 1992 → Jun 1 1992 |
Other
| Other | Proceedings of the 9th International Conference on Computational Methods in Water Resources |
|---|---|
| City | Denver, CO, USA |
| Period | 6/1/92 → 6/1/92 |