Does one filter the measured stream flow first, the filtered stream flow second, with the last pass involving filtering the secondly filtered stream flow?
Thanx Johan, been having trouble understanding that statement. Probably caught up in technical meanings.
Often, the second pass is done with time reversed (i.e. the filter applied in the negative time direction). This is to minimise phase lag in the result. This is generally a simple signal processing filter, often with a restriction put in place after each pass that the filtered flow should not exceed the observed flow. This has no physical basis, but is very simple to apply. There are other methods that attempt to have a more physical basis. The approach I developed uses a constrained deconvolution to fit the parameters. This increases the physical basis of the filter as it ensures that the effective rainfall implied by the filter is less than the estimate areal rainfall. The issue is that this requires an iterative solution. Information on this method can be found in
Conference Paper Exploring changes in catchment response characteristics: App...
The method is available in the Hydromad package (runs in R) - available from http://hydromad.catchment.org/
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India