The influence of window size on remote sensing-based prediction of forest structural variables

Ecological Processes

Table 2 Prediction results of the forest structural variables according to different window sizes (Black pine stands)

Forest variables	Window size	WorldView-3			Aerial image			LiDAR
Forest variables	Window size	R²	RMSE	RMSE%	R²	RMSE	RMSE%	R²	RMSE	RMSE%
N	100	0.78	109.37	11.40	0.80	103.59	10.80	0.84	94.75	9.87
	200	0.73	120.92	12.60	0.80	105.55	11.00	0.84	93.23	9.72
	400	0.79	108.45	11.30	0.84	93.90	9.79	0.83	96.97	10.11
	600	0.80	103.79	10.82	0.86	87.72	9.14	0.83	95.60	9.96
	800	0.82	100.33	10.46	0.88	82.68	8.62	0.84	94.79	9.88
	1000	0.86	87.24	9.09	0.83	95.31	9.93	0.84	94.57	9.86
V	100	0.79	30.50	13.46	0.77	31.30	13.81	0.76	32.62	14.39
	200	0.75	33.27	14.68	0.75	33.28	14.69	0.78	31.21	13.77
	400	0.74	33.84	14.93	0.74	33.97	14.99	0.79	30.51	13.46
	600	0.77	32.24	14.23	0.78	31.36	13.84	0.79	30.83	13.60
	800	0.77	31.78	14.02	0.75	33.53	14.80	0.79	31.10	13.72
	1000	0.78	31.63	13.95	0.76	32.53	14.36	0.77	31.72	13.99
BA	100	0.87	2.19	5.78	0.82	2.64	6.97	0.80	2.78	7.33
	200	0.82	2.62	6.92	0.76	3.01	7.94	0.81	2.68	7.08
	400	0.81	2.71	7.15	0.79	2.85	7.53	0.80	2.74	7.23
	600	0.79	2.80	7.38	0.81	2.67	7.05	0.81	2.64	6.97
	800	0.76	3.04	8.03	0.82	2.63	6.94	0.81	2.73	7.19
	1000	0.78	2.88	7.59	0.86	2.31	6.11	0.81	2.71	7.15
H	100	0.74	1.16	9.66	0.70	1.24	10.33	0.83	0.94	7.82
	200	0.73	1.17	9.76	0.74	1.16	9.68	0.83	0.94	7.83
	400	0.71	1.23	10.27	0.76	1.12	9.36	0.82	0.97	8.14
	600	0.75	1.14	9.53	0.76	1.10	9.19	0.80	1.02	8.55
	800	0.74	1.16	9.70	0.78	1.06	8.84	0.80	1.01	8.48
	1000	0.70	1.23	10.31	0.78	1.05	8.80	0.81	0.99	8.24