Single-image photogrammetry for deriving tree architectural traits in mature forest stands: a comparison with terrestrial laser scanning

Table 3 Summary of the modelling framework for building linear mixed models (LMMs) in this study; Trait_x_TLS is the TLS-derived response variable (here CPA_TLS); Trait_xn_SIP are the explanatory SIP-derived variables

No. of variables	Model ID	Formula¹
1	LMM[1a]	trait_x_TLS ~ trait_x1_SIP + (1\|stand)
1	LMM[1b]	trait_x_TLS ~ 1 + (trait_x1_SIP\|stand)
2	LMM[2a]	trait_x_TLS ~ trait_x1_SIP + trait_x2_SIP + (1\|stand)
	LMM[2b]	trait_x_TLS ~ (trait_x1_SIP\|stand) + trait_x2_SIP
	LMM[2c]	trait_x_TLS ~ trait_x1_SIP + (trait_x2_SIP\|stand)
3	LMM[3a]	trait_x_TLS ~ trait_x1_SIP + trait_x2_SIP + trait_x3_SIP + (1\|stand)
	LMM[3b]	trait_x_TLS ~ (trait_x1_SIP\|stand) + trait_x2_SIP + trait_x3_SIP
	LMM[3c]	trait_x_TLS ~ trait_x1_SIP + (trait_x2_SIP\|stand) + trait_x3_SIP
	LMM[3d]	trait_x_TLS ~ trait_x1_SIP + trait_x2_SIP + (trait_x3_SIP\|stand)

¹R notation was used; the models accounted for unexplained deviance by either fixed or random (stand) effects, assumed to be normally distributed, with mean equal to zero

ISSN: 1297-966X