![]() ![]() In plain terms, we ask a neural network to consecutively predict the weather in the next time step when showing it the weather of today. Any of these vectors are updated by a learned function, which maps the current state to the next state in the sequence. Thus, we strive to establish a formulation that remains equivariant under rotations.Ĭurrent ML-based weather prediction models treat the state of the atmosphere as a discrete series of vectors representing physical quantities of interest at various spatial locations over time. In the context of physical systems on the sphere, changes in the frame of reference are accomplished through rotations. We further expect underlying physical laws to remain unchanged if the frame of reference is altered.We do not expect physics to depend on the frame of reference.Physical laws are typically formulated from symmetry considerations: Surface windspeed predictions with SFNO and ground truth data are compared to each other.Ī potential approach to creating principled and trustworthy models involves formulating them in a manner akin to the formulation of physical laws. For more information about the math, see the ICML paper, Spherical Fourier Neural Operators: Learning Stable Dynamics on the Sphere. In this post, we discuss spherical Fourier neural operators (SFNOs), physical systems on the sphere, the importance of symmetries, and how SFNOs are implemented using the spherical harmonic transform (SHT). How can we further increase their skill, trustworthiness, and explainability, if they are not formulated from first principles?.However, these methods are purely data-driven, and you may rightfully ask: ML-based approaches do not come with such restrictions, and their uniform memory access patterns are ideally suited for GPUs. Traditional methods are formulated from first principles and typically require a timestep restriction to guarantee the accuracy of the underlying numerical method. Models such as NVIDIA FourCastNet have demonstrated that the computational time for generating weather forecasts can be reduced from hours to mere seconds, a significant improvement to current NWP-based workflows. Machine learning-based weather prediction has emerged as a promising complement to traditional numerical weather prediction (NWP) models. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |