The conceptual model that has been the focus of so much attention in the Ruataniwha dam application came under scrutiny at a board of inquiry hearing.
Niwa principal scientist Kit Rutherford spent much of yesterday explaining how his Tukituki River Model (Trim) could predict the likely effects on water quality of the river if the dam was built.
Much has been made of the model, with critics arguing it is not fit for purpose and cannot accurately predict the amount of nitrogen and phosphorus entering the river as a result of farm intensification in the wake of the irrigation scheme.
Rutherford predicts a 30 per cent increase in the loss of nitrogen from the land and a 10 per cent increase in the loss of phosphorus.
The model was developed specifically for the Tukituki, but is based on Overseer, a widely used model for predicting nutrient losses from land.
Critics point out that Overseer has an accuracy of plus or minus 30 per cent, and that this variation is consequently carried into Trim.
They say Trim modelling is accurate in only four of 12 sites being observed at present, and that the model does not consider the effect of major floods.
In his opening submission last week, Trevor Robinson, lawyer for Hawke's Bay Regional Council and its investment arm, said submitters had "significantly overstated" the role of Trim.
It did not underpin the plan change request and resource consent applications but played a minor role in assessing whether the nutrient limits in the plan change could be met if the irrigation scheme was developed.
Regardless of what the model predicted, the council's plan would impose limits on nitrogen and targets for phosphorus in order to address water quality in the Tukituki.
Answering questions by board chairman Lester Chisholm yesterday, Dr Rutherford said: "I certainly would not suggest using Trim to write consents for individual farmers, I just don't think the science is there.
"I'm confident about the predicted changes in nitrogen. I'm slightly less confident about the predicted changes in phosphorus . . . In its present form I would suggest Trim be used as I suggest: as a framework.
"At some later stage when the model is more mature . . . then I'm certainly going to revisit the issue of whether the model can be used to make a quantitative prediction of periphyton biomass."
- The Dominion Post