. Author manuscript; available in PMC: 2022 Sep 7.

Published in final edited form as: Remote Sens (Basel). 2021 Apr 30;13(9):1748. doi: 10.3390/rs13091748

Table 5. Accuracy assessment between hybrid method-based ML approaches for the whole growth season.

Methods	Stats.	LAI(m²/m²)	fCover	CCC(g/m²)
RF	R²	0.77	0.82	0.81
RF	NRMSE(%)	10.59	10.59	15.06
CCF	R²	0.59	0.65	0.55
CCF	NRMSE(%)	11.59	16.58	13.40
GPR	R²	0.70	0.68	0.60
GPR	NRMSE(%)	9.80	17.58	17.26

Note: The results of ground validation obtained from a hybrid method based on the best sample size, where the bold number indicates the best results; Stats is the statistical measures; R² is the coefficient of determination; and NRMSE (%) is a normalized root mean square error.