A generalized algebraic difference approach allows an improved estimation of aboveground biomass dynamics of Cunninghamia lanceolata and Castanopsis sclerophylla forests

Table 2 Estimated regression coefficients and goodness-of-fit statistics of Eq. 8

Stand	Coefficient	Estimate	SE	p value	Number^a	RMSE (cm)	Adjusted R ²
C. lanceolata	b ₁	0.0575	0.00340	<0.001	4534	1.810	0.935
	b ₂	0.858	0.0300	<0.001
	b ₃	0.441	0.0888	<0.001
	ρ ₁	1.581	0.0626	<0.001
C. sclerophylla	b ₁	0.0153	0.000212	<0.001	30,142	1.308	0.956
	b ₂	0.261	0.0244	<0.001
	b ₃	2.805	0.0879	<0.001
	ρ ₁	1.670	0.00661	<0.001

ρ ₁ is the autoregressive parameter; SE is the standard error of the estimate; RMSE is the root mean square error and adjusted R ² is the adjusted coefficient of determination
^aAll the possible growth intervals were used; therefore, the sample size was expanded to 4534 for C. lanceolata and 30,142 for C. sclerophylla

ISSN: 1297-966X