A key flood gauge may have underestimated the Brisbane River's flows by one-fifth, according to a report into the January flood disaster.

The Joint Flood Taskforce Report was yesterday released by Lord Mayor Campbell Newman and flood specialist Professor Colin Apelt.

The report says there was a "strong suspicion that the ... gauging station at Savages Crossing [owned by the Department of Environment and Resource Management] is seriously inaccurate causing underestimates of the flows of the order of 20 per cent or more".

Professor Apelt said the inaccuracy appeared to be caused by specialists having to extrapolate the flow rates using a statistical curve generated from "moderate" floods.

"We have questions about the accuracy of the rating curve at Savages Crossing, which is a key source of information," he said.

"At most of our gauging stations the measurements have only been taken up by 'moderate' flood levels and the rating curve has to be extrapolated upwards, which is an estimation.

"There is a strong suspicion that the extrapolation there is giving low estimates of the flow rates."

The problem is mentioned three times in the report, which also noted January's flood was higher at Savages Crossing than the 1974 event.

It estimates that the flood height at Savages Crossing was 24.167 metres at 3.40am on January 12, above the 1974 Savages Crossing flood level of 23.767 metres.

DERM director-general John Bradley last night told brisbanetimes.com.au the Savages Crossing flood gauge only measured accurately to 15.87 metres.

“So we are talking something like eight metres over the historical data,” he said.

Mr Bradfley said he could not tell whether the estimate was 20 per cent lower than the actual level, as Professor Apelt’s study suggested.

“It is quite probable that there is a variation in our predictive capacity,” he said.

“The direct consequences would be in the [flood] volume at the gauge and in the predictive inundation that occurred downstream of that gauge.”

Meanwhile, in a separate study, Seqwater found that 14 of 75 rain stations and 31 of 71 river height stations in the Brisbane River were not operating correctly in January, although it is unclear when the gauges failed.

The dam operator said that result was good considering the rain gauge network had never been tested under similar circumstances before.

“Some stations were completely destroyed by the volume and magnitude of the flood flows,” the report says.

“Certainly, any data omissions or errors resulting from these failures did not adversely impact operational decision-making.”