Results

2d Grid Comparison

• CNN-LSTM lower KGE than GLM due to lower average and standard deviation than observations.
• CNN-LSTM performed better than GLM for other metrics.
• Higher levels of bias occur for both models during the summer months. CNN-LSTM also exhibited higher errors during early Autumn.

Taylor diagram for comparison of universal (2d grid) models.

Back

Next

Deep Dive: Chapter 4.

---

Downscaling Global Climate Models with Convolutional and Long-Short-Term Memory Networks for Solar Energy Applications by C.P. Davey is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.