We are using RORB to conduct a Monte Carlo analysis. We have a concern with the application of Areal reductions factors differing in Monte Carlo and ensemble runs.
In conducting checks on the output representative hydrograph and associated storm files against the parameter tolerances specified in the release notes, we found that the rainfall applied and the resultant flow were both outside of the tolerance of 5% where ARF’s were applied within RORB, resulting in higher than expected rainfall and flows.
Where we modified the IFD table to incorporate areally reduced rainfalls, with no ARF applied by RORB, this resulted in applied rainfalls within the 5% tolerance and flows (Monte Carlo peaks and representative hydrograph peaks) 100m3/s (15%) lower.
The applied rainfall was consistently ~95% of the point rainfall depth, where the ARF should have been ~0.90. When run as an ensemble, the correct ARF appears to be applied with mean estimates consistent with the Monte Carlo estimates with the ARF applied outside of RORB.
Has anyone else encountered these issues and have a solution?
I have had a look into this and there seems to be a bug in RORB when doing embedded burst filtering while having RORB apply the ARFs in the Monte Carlo analysis. We are hoping to get this fixed soon and will release a new version of RORB when it is fixed.