DYRESM-CAEDYM - Ohau Channel diversion, Lake Rotoiti

The fate of water entering Lake Rotoiti from Lake Rotorua via the Ohau Channel may have important implications for water quality in Lake Rotoiti. 

This case study outlines some of the work undertaken by NIWA and the University of Waikato to investigate a partial or complete diversion of the Ohau Channel (which flows between Lake Rotorua and Lake Rotoiti in the central north island, New Zealand) towards the Kaituna River.

Lake Rotorua water is enriched in nutrients and organic matter, and this enrichment has the potential to add significantly to the nutrient load of Lake Rotoiti, depending on the fate and transport path of the inflow. If the Ohau Channel inflow plume entering Lake Rotoiti is warmer than water where it enters the lake, it will be buoyant and will “overflow” to become part of the upper water column (epilimnion) in the lake.  Under these circumstances nutrient inputs from Lake Rotorua may mostly diverted down the Kaituna River, with very little contribution to the eastern basin of Rotoiti.  If the inflow from the Ohau Channel is  cooler than Lake Rotoiti, it will initially “underflow” and may become either an interflow, whereby it intrudes horizontally at some intermediate depth, or be maintained as an underflow at the bottom of the water column in Lake Rotoiti.

From the point of view of water quality assessments of the Ohau Channel diversion, the main focus of this investigation is on the modelling component and use of various scenarios relating to different diversion options. The model results presented here utilise a one-dimensional model (DYRESM) of vertical stratification in Lake Rotoiti to determine the influence of the diversion on temperature and water quality properties and the ecological model CAEDYM is used to quantify the ecological effects.  The diversion options considered involve 100, 50, 10, 5 and 0% of the current Ohau Channel inflow.  

Field data relating to the volume, temperature and composition of the Ohau Channel inflow are used together with the coupled DYRESM-CAEDYM model, to more precisely define the impacts of the Ohau Channel on concentrations and distributions of dissolved oxygen, chlorophyll and nutrients in Lake Rotoiti. Relatively long-term simulations (several years) are possible with this one-dimensional approach, in contrast to the three-dimensional modelling undertaken by NIWA, which is used to simulate short-term (c. 1-4 week) events, and to examine specific design criteria for the implementation of a diversion wall or barrier.

Updated bathymetric data were required as part of this project, as an input to the three-dimensional hydrodynamic model of Lake Rotoiti and for any engineering works to be undertaken in relation to a diversion of the Ohau Channel.  The complex multi-basin shape of Lake Rotoiti, and the horizontal variability associated with the Ohau Channel plume insertion mean that there are likely to be significant horizontal variations in water properties in the lake.  This will weaken the accuracy of simulation models such as DYRESM-CAEDYM that are reliant on the dominance of vertical stratification in hindering redistribution and mixing of water, solutes and particulates in the lake. 


In this study the approach was to run DYRESM-CAEDYM over a 2-year period, with calibration of several ecological parameters until reasonable confidence in the simulation output was developed through comparisons with measured data.  In addition, a detailed field study was used to examine the dynamics of the Ohau Channel plume, and to develop a good understanding of spatial variability of water quality properties through Lake Rotoiti.  This field study was carried out from February to May 2004 and a subset of these data has also been used by NIWA to assess the performance of their three-dimensional model simulations.

Main inflows and the outflows into Lake Rotoiti were included in the application of DYRESM-CAEDYM to this project.  In addition to the specification of daily discharge required for each inflow by DYRESM, properties of the inflows are required as an input to the CAEDYM model.  These properties include concentrations of dissolved oxygen, chlorophyll a divided into buoyant and non-buoyant groups, nutrients (NO3-N, NH4-N,total nitrogen (TN), DRP and total phosphorus (TP)), biochemical oxygen demand, and suspended solids, as well as pH.  

The outflow used in the application to Lake Rotoiti is represented solely by the Kaituna River, for which only daily discharges are specified, as outflow properties are determined as part of the DYRESM-CAEDYM model simulation output.

Another forcing data input to the model is daily meteorological data.  These data include shortwave radiation, cloud cover (which is converted to longwave radiation using simulated surface water temperature as part of the calculation), air temperature, vapour  pressure (converted from relative humidity and air temperature), wind speed and

The relevant files include lake hyposography (water depth vs area), physical parameters for DYRESM, ecological parameters for CAEDYM, and an initial depth profile of temperature and water quality properties .  An input file specifying the geothermal heat flux to the bottom waters of the lake is also used on this study.


Management scenarios have been run as various reductions of the current Ohau Channel inflow (50, 10, 5 and 0%) of current inflow.  Time series plots for 0m and 60m depth are again used to make direct comparisons between the different model runs, which are for a period of 4.5 years from 1 July. The duration of the simulations was extended to 4.5 years commencing 1 July 2001 using measured data until 5 January 2004.  For the remaining 2 years the input data for 6 January 2002 to 1 January 2004 was repeated, to extend the simulation to 31 December 2005.  For most variables there is little purpose in carrying out quantitative comparisons between the different flow scenarios  but where warranted (e.g. important variables where differences are clearly evident in the simulations) a quantitative comparison has been carried out. 



Changes in water quality of the Kaituna River outflow are likely to occur with a diversion.  The assessments in this study have considered a 100% diversion and the present case, i.e., no diversion.  

Until more information is obtained on the rate of sedimentation of suspended particulates associated with the Ohau Channel inflow, it is likely that our results overestimate the concentrations of suspended solids and particulate nutrients associated with Ohau Channel water that is diverted directly towards the outflow.  For the diversion case there are likely to be slightly higher levels of nutrients, chlorophyll a and suspended solids in the Kaituna River, reflecting an increasing proportion of water received directly via the Ohau Channel than currently exists.

Associated Models

Dynamic Reservoir Simulation Model - Computational Aquatic Ecosystem Dynamics Model (DYRESM- CAEDYM)

DYRESM – CAEDYM is a one dimensional hydrodynamic-ecological model that can be used to investigate the interactions between physical, chemical and biological processes that occur in lakes and reservoirs over time scales ranging from days to seasonal to interannual.



 Lake Rotoiti - Fieldwork and modelling to support considerations of Ohau Channel diversion from Lake Rotoiti. (2005): Prepared by  David Hamilton, Chris McBride & Toradji Uraoka Centre for Biodiversity and Ecology Research, University of Waikato, New Zealand