Table 4. Performance of Single Model (SM), Modified Single Model (MSM), Group Model (GM), and Local Model (LM) by region for calibration and validation dataset. (a) Root mean square error (RMSE), (b) Bias, (c) Pearson’s correlation coefficient (r).

(a)
Region	Model
	SM		MSM		GM		LM
	RMSE_cal	RMSE_val	RMSE_cal	RMSE_val	RMSE_cal	RMSE_val	RMSE_cal	RMSE_val
Southwest	4.88	4.06	3.32	4.60	2.79	4.64	2.14	4.76
Southeast	6.16	6.62	3.27	6.12	3.89	6.15	2.23	5.48
North central	6.21	5.09	3.80	3.49	5.69	4.57	3.06	4.40
Seoul	6.49	7.05	4.35	4.63	6.09	6.20	3.64	5.63
Total	6.13	6.11	3.90	4.62	5.25	5.57	3.09	5.17
(b)
Region	Model
	SM		MSM		GM		LM
	Bias_cal	Bias_val	Bias_cal	Bias_val	Bias_cal	Bias_val	Biascal	Bias_val
Southwest	-3.32	-0.95	-7.31×10^-14	2.37	-0.43	0.95	-0.14	1.70
Southeast	-4.75	-2.60	-1.52×10^-14	2.15	-0.04	1.55	0.11	1.35
North central	1.79	2.06	1.42×10^-14	0.27	0.48	0.69	-0.43	-0.20
Seoul	1.81	4.53	-9.74×10^-14	2.70	-0.24	1.78	-0.20	2.60
Total	-0.10	1.92	-1.93×10^-14	1.91	-0.06	1.32	-0.20	1.52
(c)
Region	Model
	SM		MSM		GM		LM
	r_cal	r_val	r_cal	r_val	r_cal	r_val	r_cal	r_val
Southwest	0.62	0.51	0.69	0.54	0.72	0.56	0.85	0.56
Southeast	0.76	0.19	0.79	0.15	0.75	0.21	0.87	0.25
North central	0.60	0.79	0.83	0.89	0.64	0.80	0.89	0.82
Seoul	0.67	0.70	0.83	0.86	0.68	0.63	0.86	0.76
Total	0.70	0.74	0.83	0.82	0.71	0.72	0.88	0.77