Gaps in the research

  • Few studies focused on downscaling global solar radiation at key solar sites especially within the Queensland region.
  • Reanalysis products more common rather than direct use of GCM models - both approaches introduce biases.
  • Lack of application of deep learning methods to downscaling for global solar radiation, despite its promising application to other climate variables.
  • Lack of recommendations for the construction of end to end deep learning architecture for the downscaling task, including evaluation of configuration and parameters for respective modules.
  • Few studies investigate uncertainty of downscaling models other than review of evaluation metrics.
  • Few studies perform future projection via resulting models to evaluate differences under climate forcing.

    • Creative Commons License
    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.