A prominent statistician once said that all computer models are wrong, but some are useful.
In the face of this faint praise, one academic group claims to have developed models that work well enough to help planners forecast beach erosion accurately in areas like the Cooloola Coast.
A University of New South Wales study has found shoreline prediction models claimed to be effective at forecasting changes to natural, sandy beaches with an accuracy of approximately 10m.
A spokesperson said accurate predictions would help governments, planners and communities to make critical decisions about coastal development, risk management and environment protection.
“Our results indicate that certain beaches can be modelled nearly as well as they can be remotely observed,” the study’s lead author Dr Yongjing Mao said.
Dr Mao, of the university’s School of Civil and Environmental Engineering, said further benchmarking is needed to understand model performance for coastlines in urbanised areas, where human-made structures complicate shoreline dynamics.
In the study, 34 shoreline models from various modelers around the world were evaluated on their ability to make predictions of a shoreline position as part of a blind competition, testing their predictions against observed results.
“The researchers found that the top-performing models could predict shoreline change with an accuracy of approximately 10 metres for bay-shaped beaches over both short and medium time scales,” Dr Mao said.
“We found that most shoreline models successfully capture both the response to storms in not only short-term but also medium-term predictions,” the researcher said.
.
