GNS Science's assessment of risk following the Canterbury earthquakes has been criticised as unachievable at the Royal Commission's hearings.
The Royal Commission hearings began this week with evidence from scientists about the earthquake sequence which has affected Canterbury since 4 September last year.
On Wednesday, evidence was given by Norman Abrahamson, an Adjunct Professor of Civil Engineering from the University of California, who peer reviewed GNS Science's report to the commission.
Professor Abrahamson said he believes GNS Science's suggested changes to the building code of the allowable risk of collapse in a significant quake goes too far.
He said in California, the acceptable risk levels are two to 10 times higher than those proposed by GNS and New Zealand will have to decide if it wants more safety at a higher cost, or can accept the kind of risk figures used around the world.
However, Professor Abrahamson disagreed with GNS raising the magnitude of earthquake which could be considered to be a significant risk to buildings, from magnitude 5 to 5.5, saying a particularly energetic quake measuring 5 could still cause damage to buildings.
The expert told the commission GNS Science may have over-estimated how much the ground shook during the September and February quakes.
Professor Abrahamson said the GNS report states that the quakes on 4 September, 22 February and 13 June had above average ground shaking and their assumption is that future tremors in Canterbury may have similarly high ground shaking.
However, he questioned the method used for the calculations and did not agree it is likely that future quakes will have higher than average ground movement.