CLUES - Northland Sediment: E Coli Study

This case study provides an example of the development and calibration of a customised version of the Catchment Land Use for Environmental Sustainability model (CLUES).

The CLUES model was used as a part of the Northland Sediment Study (NSS) which had the aim of assessing the potential economic costs of achieving sediment and microbial water quality objectives in the Whangarei Harbour catchment area. The study had two main objectives:

  1. Develop model frameworks and outputs that will enable the assessment of catchment sediment and E. coli loads and the expression of the environmental outcomes of these loads as attributes. MPI has contracted NIWA to deliver this objective.
  2. Incorporate the model frameworks and outputs developed in Objective 1 into a catchment economic model that will be used to identify cost-effective ways to manage sediment and E. coli loads in the Whangarei Harbour catchment. MPI is contracting another provider to deliver this objective.

The CLUES model was used to address the first objective of the NSS, being a key component of the assessment sediment and more importantly E. coli loads throughout the catchment and wider region. This proportion of the study then provided inputs to an economic model (NZFARM developed by Dr Adam Daigneault of Landcare Research) which was designed to address the financial implications of adopting various mitigation strategies within the harbour catchment area. 


The CLUES model was used to determine mean annual loads of E. coli throughout northland waterways with particular focus on Whangarei harbour. The catchment of interest is broken into subcatchments, and each subcatchment has a number of landuses with associated yields, which are modified according to environmental factors such as rainfall. These sources are accumulated and decayed down the stream network, with addition of point source loading.

The CLUES model was identified as a suitable model for this case study but was customised specifically for this application, changes included:

  • Introduction of a slope term.
  • Adjustment of model parameters where appropriate to improve the fit of model predictions with measured loads.
  • Use of a formal method for assessing goodness-of-fit and the variability and interdependency of the adjusted model parameters.

Calibration of the model to suit the Northland focus area was carried out using twenty-five water quality sites, five of these were located in within the Whangarei Harbour catchment. Data at these key river drainage sites for current E. coli loads was used to calibrate the clues model in a regional context. In addition to this, MPI and Northland Regional Council required the current E. coli loads and concentrations at 11 key sites in the harbour catchment to be estimated.

CLUES nodes of importance 11


 CLUES calibration northland

Northland region showing streams of order ≥ 3, lakes, major towns, the Whangarei Harbour catchment and the calibration sites


 Findings and Conclusions

The CLUES generated E. coli load predictions were converted to E. coli concentrations under the assumption of a linear relationship between load and concentration. This was useful in terms of the planning of mitigation strategies for which if a mitigation option reduced loads by X% then concentrations were also to be assumed to drop by X%.

Model predictions in some cases were counter-intuitive when land use was considered. For example, selected forested catchments in Northland exhibited E. coli concentrations that were larger than those predicted for catchments where pastoral land use was dominant. While these apparent anomalies cannot currently be explained possible inputs include waste from feral animals.

Within the Whangarei Harbour catchment, the model predicted that the overwhelming bulk of the E. coli load was derived from streams flowing directly into the harbour, rather than the pseudocatchment or point source discharges.     

Generally there seems to be large uncertainties in the CLUES model predictions which largely come from the amount and significance of unknown factors. These include; lack of water quality monitoring sites, lack of knowledge around E. coli land and stream dynamics, lack of ground water modelling and uncertainties in the measurement of measured loads that were used for calibration. This level of uncertainty is something that is noted in this case study, which highlights the importance of acknowledging this uncertainty when considering future planning and management options.

Associated Models

Catchment Land Use for Environmental Sustainability (CLUES)

CLUES is a modelling system for assessing the effects of land use change on water quality and socio-economic factors at a minimum scale of sub-catchments (~10 km2 and above).


Christopher Palliser, Sandy Elliott, and Sharleen Yalden (2015): Northland Sediment Study - E. coli modelling. Prepared for Ministry for Primary Industries. NIWA Client Report No: HAM2015-122