University of Twente Student Theses
Numerical vs AI Models in Global Hurricane Forecasting
Nessipbayev, A. (2024) Numerical vs AI Models in Global Hurricane Forecasting.
|
PDF
874kB |
| Abstract: | Natural disasters, such as hurricanes and floods are often recognized as catastrophic events that have a major impact on socio-economic and envi- ronmental sectors. They are considered to be difficult to control with the current level of technology that humans have. Thus, it is necessary to have a good prediction mechanism that can provide an early alert to save as many lives as possible and reduce damage. Artificial intelligence (AI) is consid- ered to be one of the most promising solutions to potential problems that humans may face as a civilization. It is not uncommon for agencies such as the National Centers for Environmental Prediction (U.S.) or the European Centre for Medium-Range Weather Forecasts (ECMWF) to develop and de- ploy different types of AI technologies that include machine learning and neural networking to analyze a large amount of data received via satellites or on-ground sensors. With the help of AI, the agencies are looking for po- tential improvements in the accuracy of prediction compared to times when most things were calculated by complex machines that required significant time and resources. This research specifically focuses on performance comparison of the modern numerical models that are used for accurate weather prediction and their AI-enhanced counterparts. The findings highlight the potential of AI-enhanced models to improve the prediction of hurricanes, which ultimately leads to better preparation and more efficient efforts of local communities. Moreover, the work provides insights into model performances and demonstrates that ,even though the AI-system provide acceptable and positive results, they do not show crucial difference in prediction accuracy to state that they are significantly better than the deployed numerical systems. Lastly, the summary of the findings gives a following result: The IFS took the 1st place, with the GFS Graphcast in 2nd, closely followed by AIFS in 3rd, and the GFS as the last one on the list. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 01 general works, 31 mathematics, 54 computer science |
| Programme: | Business & IT BSc (56066) |
| Link to this item: | https://purl.utwente.nl/essays/101166 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page