An Assessment of GCM Performance at a Regional Scale Using a Score-Based Method

Table 3

Model performances for precipitation.

GCMs	GCM number	Mean Value (mm/y)	StDev	NRMSE	Monthly corr.	Spatial corr.	M-K		EOF		PDF		Total RS
							Zc	Slope (mm/10 a)	EOF1	EOF2	BS	Sscore
obs	0	568.2	61.5				−0.86	−11.76	71.72	8.69
access1-0	1	775.0	62.3	0.72	0.95	0.82	−0.34	−2.63	53.21	17.70	0.03	91.22	13.22
bcc-csm1	2	955. 3	67.5	1.00	0.92	0.61	0.12	2.65	61.31	14.08	0.06	90.13	27.21
bnu-esm	3	1,255.9	83.4	1.50	0.94	0.45	0.12	1.78	60.56	16.00	0.10	81.30	48.52
canesm2	4	914.3	75.5	1.04	0.85	0.72	1.46	21.48	49.79	21.16	0.04	80.24	39.59
ccsm4	5	891.0	65.6	0.92	0.93	0.70	−1.10	−13.43	60.89	13.03	0.03	81.26	18.98
cesm1-bgc	6	888.2	50.7	0.90	0.93	0.70	−2.30	−19.45	58.82	13.27	0.03	81.75	23.99
cnrm-cm5	7	703.6	71.8	0.65	0.96	0.80	1.21	14.66	55.57	16.66	0.03	81.85	19.68
giss-e2-h	8	802.2	77.5	0.80	0.91	0.57	0.23	2.04	64.49	15.18	0.04	93.24	26.15
csiro-mk3.6	9	650.6	51.7	0.54	0.98	0.75	1.08	10.45	62.30	10.64	0.01	89.64	12.24
fgoals-g2	10	881.0	60.3	1.01	0.78	0.81	−0.59	−6.11	46.99	25.29	0.06	84.18	30.79
gfdl-cm3	11	879.7	65.72	0.88	0.92	0.52	−1.13	−16.57	57.97	19.90	0.05	85.45	27.14
hadgem2	12	725.4	64.31	0.70	0.91	0.76	1.16	9.07	49.51	18.78	0.02	82.22	21.85
inm-cm4	13	923.5	62.61	0.90	0.95	0.68	1.30	13.30	62.19	13.17	0.05	89.46	23.00
noresm1-m	14	1,035.7	65.13	1.12	0.94	0.54	−0.80	−5.93	70.20	10.14	0.04	74.48	26.06
miroc-esm	15	898.3	68.81	1.04	0.83	0.57	−0.18	−1.78	65.68	13.81	0.03	76.61	30.46
mpi-esm-lr	16	887.8	63.13	0.95	0.89	0.66	0.45	3.98	55.38	18.05	0.04	89.27	27.01
ipsl-cm5b-lr	17	674.1	66.00	0.71	0.90	0.71	2.39	22.50	49.97	14.25	0.02	73.81	27.14
mri-cgcm3	18	456.1	33.58	0.72	0.93	0.76	−0.15	−1.38	48.21	19.47	0.01	61.28	28.21

EOFs 1 and 2 are percentages of the explained variance for the first two leading modes of each EOF; PDF is the probability density function.