Matthew Keir
posted this
05 October 2021
- Last edited 05 October 2021
Hi Ziming,
Tributary factors are derived from catchment areas. Clearly, this is an approximation and assumes in the long run rainfall is equally likely across the catchments and that the temporal aspect of river flow is not material to the type of modelling the dataset is used for.
There are two tributary flow files in the dataset for the Waikato system, total trib flows to Arapuni and to Kapapiro. We treat these as two observations (with their inherent measurement error) that we fit to our model based on catchment size.
Using Arapuni as the basis for the tributary factors (as this is the longest measurement series), trib flows to Karapiro are 1.2318638 times that to Arapuni.
After 1947 we've derived an Arapuni series from the two observations [(ArapuniTrib + (KarapiroTrib/1.2318638))/2].
Prior to 1947, there is no Karapiro series, so the Arapuni series in the dataset is used.
Trib factors based on catchment areas are then applied to derive tributary flows for each station in the Waikato hydro chain.
This method explains why you are observing differences in the derived series we publish and what you calculate when you only consider only the Arapuni trib file in the dataset. We've provided these derived tributary flow series to make it easier for our users to get the direct input data they need rather than everyone needing to derive these files themselves.
I hope that helps.
Cheers,
Matthew