September 25th 2015

Post date: 25-Sep-2015 06:55:36

I've updated the model testing page with the second major evolution of the IoW model. This additional training data has enabled better predictions to be made due to the wider geographic spread in the training data. This data incidentally is all based on lure/harp trap identification which was carried out by Ian Davidson-Watts in the period 2006-2007, and highlights the need for both acoustic and more the more intensive on site trap based surveys. Next step is to start to examine the detailed contributions from the current environment grids and look to see if what is driving the model performance.