Machine Learning as a Downscaling Approach for Prediction of Wind Characteristics under Future Climate Change Scenarios

Table 2

Comparison of statistical predictive measures across all the tested scenarios.

Model	Training (70/30)		Verification
	CC	RMSE	Correlation coefficient			Root mean square error
			65/35	70/30	75/25	65/35	70/30	75/25
A	0.9028	2.0117	0.3642	0.1237	0.2136	5.7777	17.4996	4.1732
Swinter	0.8944	2.0585	0.0514	0.2003	0.1699	32.723	10.9887	12.8441
Sspring	0.0896	1.8625	0.1296	0.1679	0.6327	15.2014	10.8614	3.0635
Ssummer	0.8934	1.8063	0.0744	0.2707	0.1093	30.627	6.2338	15.2301
Sfall	0.888	2.0465	0.075	0.5276	0.5049	3.9912	3.6316	3.8453
Mean	0.6914	1.9434	0.0826	0.2916	0.3542	20.6356	7.9289	8.7458
MJan	0.8983	2.0922	0.56	0.8054	0.7704	3.7827	2.6749	2.8072
MFeb	0.9046	1.9672	0.5936	0.7791	0.0881	3.6361	2.6831	41.9419
MMar	0.9026	1.9325	0.7804	0.7974	0.6843	2.6535	2.5707	3.0753
MApr	0.8988	1.8653	0.1763	0.7333	0.4716	12.1353	2.6809	4.2891
MMay	0.897	1.7869	0.2977	0.6998	0.0389	5.8888	2.6363	47.1423
MJun	0.8892	1.7279	0.558	0.6979	0.3844	3.0569	2.4745	4.5731
MJul	0.8822	1.6457	0.5116	0.6576	0.6059	2.9252	2.6373	2.6362
MAug	0.8903	1.6815	0.7566	0.7566	0.7392	2.2544	2.2544	2.2925
MSep	0.902	1.9198	0.3506	0.7627	0.0679	5.77	2.6672	40.5325
MOct	0.8815	1.9988	0.0362	0.7565	0.4504	3.86	2.5783	4.0723
MNov	0.889	1.9712	0.0423	0.7174	0.0614	73.324	2.7225	22.6081
MDec	0.8984	2.0211	0.3008	0.7511	0.0912	7.0636	2.7503	24.5093
Mean	0.8945	1.8842	0.4137	0.7429	0.3711	10.5292	2.6109	16.7067