Convolutional conditional neural processes for local climate downscaling
| Autor: | J. Scott Hosking, Anna Vaughan, Richard E. Turner, William Tebbutt |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2022 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning J.2 010504 meteorology & atmospheric sciences Computer science 0207 environmental engineering FOS: Physical sciences 02 engineering and technology Machine learning computer.software_genre 01 natural sciences Machine Learning (cs.LG) Set (abstract data type) symbols.namesake Climate impact 020701 environmental engineering Representation (mathematics) Gaussian process Physics::Atmospheric and Oceanic Physics 0105 earth and related environmental sciences QE1-996.5 Training set business.industry Deep learning Geology Physics - Atmospheric and Oceanic Physics 13. Climate action Atmospheric and Oceanic Physics (physics.ao-ph) symbols Artificial intelligence business computer Downscaling |
| Zdroj: | Geoscientific Model Development, Vol 15, Pp 251-268 (2022) |
| ISSN: | 1991-9603 |
| Popis: | A new model is presented for multisite statistical downscaling of temperature and precipitation using convolutional conditional neural processes (convCNPs). ConvCNPs are a recently developed class of models that allow deep learning techniques to be applied to off-the-grid spatio-temporal data. This model has a substantial advantage over existing downscaling methods in that the trained model can be used to generate multisite predictions at an arbitrary set of locations, regardless of the availability of training data. The convCNP model is shown to outperform an ensemble of existing downscaling techniques over Europe for both temperature and precipitation taken from the VALUE intercomparison project. The model also outperforms an approach that uses Gaussian processes to interpolate single-site downscaling models at unseen locations. Importantly, substantial improvement is seen in the representation of extreme precipitation events. These results indicate that the convCNP is a robust downscaling model suitable for generating localised projections for use in climate impact studies, and motivates further research into applications of deep learning techniques in statistical downscaling. 26 pages, 12 figures |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
| Nepřihlášeným uživatelům se plný text nezobrazuje | K zobrazení výsledku je třeba se přihlásit. |