Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Monthly Diffusion v0.9 (MDv0.9) is introduced as a climate emulator employing a Conditional Variational Auto-Encoder (CVAE) architecture inspired by spherical Fourier neural operators (SFNO) to model low-frequency atmospheric variability. It operates at a 1.5-degree grid spacing and monthly mean timesteps, targeting data-sparse regimes with limited computational resources. Initial results and the architectural design and training procedure are detailed.
Climate emulation gets a boost with MDv0.9, a latent diffusion model that achieves meaningful results even with sparse data and modest compute.
Here, we describe Monthly Diffusion at 1.5-degree grid spacing (MD-1.5 version 0.9), a climate emulator that leverages a spherical Fourier neural operator (SFNO)-inspired Conditional Variational Auto-Encoder (CVAE) architecture to model the evolution of low-frequency internal atmospheric variability using latent diffusion. MDv0.9 was designed to forward-step at monthly mean timesteps in a data-sparse regime, using modest computational requirements. This work describes the motivation behind the architecture design, the MDv0.9 training procedure, and initial results.