Table 2.
Validation Statistics.
| Tmax | Tmin | RHave | VPD | WBGT | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Cor | MAE | Cor | MAE | Cor | MAE | Cor | MAE | Cor | MAE | |
| Global | 0.86 | 1.1 | 0.73 | 1.3 | 0.76 | 4.83 | 0.76 | 0.15 | 0.86 | 0.77 |
| Africa | 0.80 | 1.07 | 0.67 | 1.28 | 0.75 | 4.78 | 0.76 | 0.19 | 0.78 | 0.72 |
| Asia | 0.89 | 0.95 | 0.72 | 1.15 | 0.75 | 4.44 | 0.77 | 0.15 | 0.86 | 0.71 |
| Europe | 0.95 | 0.96 | 0.84 | 1.24 | 0.8 | 4.89 | 0.82 | 0.11 | 0.95 | 0.7 |
| North America | 0.86 | 1.28 | 0.75 | 1.49 | 0.76 | 5.14 | 0.73 | 0.14 | 0.89 | 0.87 |
| Oceania | 0.92 | 1.00 | 0.73 | 1.47 | 0.83 | 4.51 | 0.8 | 0.15 | 0.91 | 0.79 |
| South America | 0.74 | 1.33 | 0.66 | 1.16 | 0.69 | 5.18 | 0.69 | 0.17 | 0.79 | 0.84 |
Table 2 depicts the median Spearman’s correlation coefficient (Cor) and the mean absolute error (MAE) for daily data by continent during the hottest 3-month period. All correlations are significant (p-value < 0.01). The CHC-CMIP6 observational products perform well for each of the five variables globally as well as regionally. The spatial distribution of correlations may be seen in the SI.