Pogoda i klimat

Weather and climate modelling · id earthsystems · aliases: ML weather prediction, AI weather forecasting, machine learning weather prediction, prognoza pogody AI, climate AI

Uczenie maszynowe w naukach o Ziemi: globalne prognozy średnioterminowe, prognozy zespołowe, prognozy krótkoterminowe i emulacja modeli klimatu.

Machine learning in the Earth sciences: global medium-range forecasts, ensemble forecasts, nowcasting and climate-model emulation.

In one sentence

Modele uczone na dziesięcioleciach zapisów atmosfery, które prognozują pogodę szybciej niż klasyczne symulacje.

Models trained on decades of atmospheric records that forecast the weather faster than classical simulations.

Problem it solves

Tradycyjna prognoza wymaga godzin pracy superkomputera, więc trudno często ją odświeżać i liczyć wiele wariantów.

A traditional forecast takes hours on a supercomputer, so it is hard to refresh often or run many variants.

What it does

Zamiast liczyć fizykę od zera, model patrzy na obecny stan atmosfery i z doświadczenia przewiduje jej stan za kilka godzin, krok po kroku.

Instead of computing physics from scratch, the model looks at the current atmosphere and, from experience, predicts its state hours ahead, step by step.

Everyday example

Służby kryzysowe dostają wcześniej ostrzeżenie o trasie huraganu, a operator farm wiatrowych planuje produkcję prądu na kolejne dni.

Emergency services get earlier warning of a hurricane's path, and a wind-farm operator plans power output for the coming days.

This concerns you if you…

• planujesz pracę zależną od pogody
• pracujesz w energetyce, rolnictwie lub zarządzaniu kryzysowym
• plan work that depends on the weather
• work in energy, farming or emergency management

Subtopics

• Globalne prognozy pogody z sieci neuronowych (Data-driven global medium-range weather forecasting): Deterministic models (GNNs, transformers) trained on reanalysis that autoregressively forecast the global atmospheric state 1–15 days ahead.
• Modele hybrydowe i emulatory klimatu (Hybrid and learned climate models (climate emulators)): Learned emulators of general circulation models and hybrids that pair a differentiable dynamics solver with networks parameterising sub-grid processes.
• Probabilistyczne prognozy pogody (zespoły) (Probabilistic ensemble weather forecasting): Generating forecast ensembles with probabilistic (mainly diffusion) models instead of a single deterministic forecast.
• Prognozy opadów na najbliższe godziny z radaru (nowcasting) (Precipitation nowcasting from radar): Short-range (5–90 min, up to a few hours) precipitation forecasting from radar image sequences with generative or convolutional models.
• Zwiększanie szczegółowości prognoz i projekcji klimatu (downscaling) (Machine-learning downscaling of weather and climate data): Mapping coarse fields from global models or reanalysis to high resolution with regression and generative networks (e.g. residual-correcting diffusion).

Related topics

• Monitoring rolnictwa, lasów i katastrof z satelity (Satellite monitoring of agriculture, forests and disasters)
• Sieci grafowe (GNN) (Graph neural networks (GNN))

Sources

• GraphCast: Learning skillful medium-range global weather forecasting · Remi Lam, Alvaro Sanchez-Gonzalez, Matthew Willson, Peter Wirnsberger, Meire Fortunato i in. (Google DeepMind)
• GenCast: Diffusion-based ensemble forecasting for medium-range weather · Ilan Price i in. (Google DeepMind)