USGS Scientist Co-authored Paper on Predicting Ripple Wavelength
Meg Palmsten (SPCMSC) co-authored a recently published manuscript titled, “A machine learning approach to predicting equilibrium wavelength.”
Understanding sand ripples caused by waves is an important component of forecasting coastal change hazards. Ripples affect bottom boundary layer dynamics including wave attenuation and sediment transport. This manuscript describes a new ripple predictor that produces a probability distribution of ripple wavelengths and an estimate of ripple heights. The new predictor combines two new machine learning models with two widely used empirical ripple parameterizations. Testing of the predictor on independent datasets demonstrated that the new model is more accurate than each individual model. The new predictor can now be incorporated into models used to study coastal change hazards.