Table 3. Variable combinations and classification accuracy results for the different models.

Model number			#1	#2	#3	#4	#5	#6
Total number of variables in model			14	18	2	5	8	11
Model variables	Bathymetry		√	√	√	√	√	√
	Bathy derivatives	Complexity	√	√				√
		Aspect	√	√
		BPI	√	√
		Slope	√	√
		Maximum Curvature	√	√
		Rugosity	√	√				√
	Mosaic		√	√
	Mosaic derivatives	Red HSI	√	√			√	√
		Green HSI	√	√			√	√
		Blue HSI	√	√			√	√
		GLCM Homogeneity	√	√
		GLCM Entropy	√	√
		GLCM Correlation	√	√				√
	AR derivatives	AR mean		√	√	√	√	√
		AR slope		√		√	√	√
		AR skewness		√		√	√	√
		AR kurtosis		√		√	√	√
Model performance	Overall accuracy (%)		88.5	93.6	90.2	89.7	92.2	93.5
	Kappa coefficient		0.81	0.90	0.84	0.83	0.88	0.90

The tick symbol (√) indicates the variable was used in a model.