Spatio-temporal forecasting on WeatherBench-S 2017-2018 (test)
1.018Temperature MSESimVP+SVQ
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SimVP+SVQcodebook=Learnable2023.12 | 1.018 | 0.6109 | 30.611 | 3.657 | 1.4186 | 0.7858 | 0.0458 | 0.1463 | |
| SimVP+SVQcodebook=Frozen2023.12 | 1.023 | 0.6131 | 30.863 | 3.661 | 1.4337 | 0.7861 | 0.0456 | 0.1456 | |
| SimVPVQ module=None2023.12 | 1.105 | 0.6567 | 31.332 | 3.776 | 1.4996 | 0.8145 | 0.0466 | 0.1469 | |
| TAU2023.12 | 1.162 | 0.6707 | 31.831 | 3.818 | 1.5925 | 0.8426 | 0.0472 | 0.146 | |
| MAU2023.12 | 1.251 | 0.7036 | 34.529 | 4.004 | 1.9001 | 0.9194 | 0.0496 | 0.1516 | |
| PredRNN2023.12 | 1.331 | 0.7246 | 37.611 | 4.096 | 1.881 | 0.9068 | 0.055 | 0.1588 | |
| ConvLSTM2023.12 | 1.521 | 0.7949 | 35.146 | 4.012 | 1.8976 | 0.9215 | 0.0494 | 0.1542 | |
| PredRNN.V22023.12 | 1.545 | 0.7986 | 36.508 | 4.087 | 2.0072 | 0.9413 | 0.0505 | 0.1587 | |
| E3D-LSTM2023.12 | 1.592 | 0.8059 | 36.534 | 4.1 | 2.4111 | 1.0342 | 0.0573 | 0.1529 | |
| PredRNN++2023.12 | 1.634 | 0.7883 | 35.146 | 4.012 | 1.8727 | 0.9019 | 0.0547 | 0.1543 | |
| MIM2023.12 | 1.784 | 0.8716 | 36.534 | 4.1 | 3.1399 | 1.1837 | 0.0573 | 0.1529 |