Two new methods applied to crown width additive models: a case study for three tree species in Northeastern China

Table 8 The goodness-of-fit statistics of the NSURMEM and GAM systems for larch, birch, and poplar

CR	NSURMEM system				GAM system				FI%
CR	ME	MAE	RMSE	FI	ME	MAE	RMSE	FI	FI%
Larch
CRE	0.0071	0.3844	0.5075	0.3158	0.0000	0.3799	0.5009	0.3334	5.59
CRS	0.0020	0.3793	0.4971	0.3477	0.0000	0.3729	0.4898	0.3668	5.47
CRW	− 0.0017	0.3715	0.4837	0.3248	0.0000	0.3702	0.4824	0.3284	1.08
CRN	0.0014	0.3854	0.5017	0.3041	0.0000	0.3815	0.4968	0.3175	4.42
CW	0.0044	0.5014	0.6745	0.5092	0.0000	0.4883	0.6598	0.5303	4.14
Birch
CRE	0.0348	0.4175	0.5601	0.2323	0.0000	0.4194	0.5555	0.2450	5.45
CRS	0.0453	0.4048	0.5365	0.2889	0.0000	0.4065	0.5326	0.2993	3.59
CRW	0.0269	0.3803	0.5036	0.2306	0.0000	0.3806	0.5020	0.2354	2.08
CRN	0.0541	0.3686	0.4938	0.2399	0.0000	0.3695	0.4906	0.2500	4.20
CW	0.0805	0.4895	0.6637	0.4515	0.0000	0.4886	0.6559	0.4643	2.84
Poplar
CRE	0.0314	0.4030	0.5266	0.2147	0.0000	0.3762	0.4973	0.2997	39.56
CRS	− 0.0723	0.4234	0.5508	0.2745	0.0000	0.4126	0.5359	0.3134	14.15
CRW	− 0.0247	0.4204	0.5473	0.3604	0.0000	0.4164	0.5459	0.3637	0.92
CRN	0.0139	0.3803	0.5071	0.3069	0.0000	0.3803	0.5064	0.3089	0.66
CW	− 0.0259	0.4570	0.6036	0.5456	0.0000	0.4379	0.5884	0.5682	4.15

Note: The bolded numbers indicate the best performance for each case of the NSURMEM and GAM systems

ISSN: 1297-966X