Project title: Benchmarking
The component
Benchmarking is being developed as an integral part of the NZWaM-Hydro project with the aim of:
- providing objective proof of model suitability and improvements across a range of spatial and temporal scales;
- identifying issues with observations and model developments; and
- identifying locations where monitoring network effort is needed to improve hydrological processes understanding.
NZ Benchmarking framework is based on initial work by McMillan et al (2016), and is developed to integrate land surface, groundwater, hydrological and isotope signatures and statistics.
NZ Benchmarking will specifically assess the impacts of the following factors on hydrological model prediction uncertainty:
- spatial and/or temporal resolution of input data
- model structure (i.e. lumped/semi-distributed/distributed).
As part of this task, data will be acquired from Regional Councils, Crown Research Institutes, remote-sensing platforms, the NIWA Climate Database, and via collaboration with industry sectors.
NZ benchmarking will assist model development and calibration and help end-users to select the appropriate model specific to their needs based on objective performance measures.
The objective
The purpose of the benchmarking task is to develop and implement a state-of-the-art operational method (that can be applied at national scale) to test hydrological models in New Zealand.
This will be tested across a variety of surface water,groundwater and isotopic signatures (e.g. mean annual discharge, mean annual low flows,flow duration curve, mean residence time). This will require the development, set-up and operational updating of quality-assessed observed hydrological data(discharge, groundwater level, lake level, soil moisture, drainage, evaporation time series) and associated signatures.
The NZ Benchmarking procedures will be conducted on NIWA’s High Performance Computing Facility and will be implemented within HydroDesk-NZ.
The project is funded through the Strategic Science Investment Fund from the Ministry of Business, Innovation and Employment.
Planned tasks for the year
- Update of hydrological dataset (high frequency precipitation, continuous streamflow information) across three digital river networks.
- Update of augmented precipitation gridded data due to the addition of over 1038 regional council sites.
- Comparison of NZWaM-Hydrology surface water models in the Mataura catchment run with different combinations of digital river network. This will assess the impact of rainfall station density, and different soil and geological characteristics datasets on model prediction uncertainty.
Progress to date
- Updated precipitation and discharge from regional councils to be used as part of hydrological model development. Information to be provided either as Tideda file or netcdf files.
- Updated 5km resolution augmented precipitation gridded time series to December 2017.
- Refinement of the NZ Benchmarking procedures and hydrological signatures on the Mid-Mataura case study.