Table editor GitHub repo

Kilometer-Scale ML Weather Forecasting

Research map for kilometer-scale forecasting and downscaling

This page is meant to serve as a living repository of research efforts applying machine learning to kilometer-scale weather forecasting. The main view is organized first by task, then by whether the model is deterministic or probabilistic. Global models are included only as a narrowly scoped reference set when they are commonly used as drivers, priors, or baselines for kilometer-scale work.

Last updated: -- 0 papers

KM Forecasting Models

True kilometer-scale forecasting systems or regional LAMs.

Deterministic

Single best-estimate forecasting models without explicit probabilistic outputs.

Probabilistic

Forecasting models with explicit uncertainty estimates, distributions, or ensemble generation.

KM Downscaling

Regional super-resolution, statistical downscaling, and benchmark efforts delivering kilometer-scale outputs.

Deterministic

Downscaling and benchmark systems focused on single predictions or deterministic baselines.

Probabilistic

Downscaling systems that represent uncertainty explicitly, including ensemble and distributional methods.

Selected Global Context

Intentionally limited to a small set of widely referenced global drivers, priors, or baselines relevant to kilometer-scale modeling.