23 papers in the archive.
A CNN generates dynamically consistent counterfactual extreme heat events, estimating the 2023 southcentral North America heat was 1.18–1.42°C warmer due to warming.
Climate analysis & attributionColorado State University (incl. CIRA)Stanford University
Review of ML for climate physics and climate simulations, arguing physics knowledge and interpretability are essential in the small-data, non-stationary regime of long-term climate projection.
Hybrid physics-ML & parameterizationMassachusetts Institute of Technology (MIT)Stanford UniversityUniversity of Chicago
Perspective on climate model hierarchies since 2016, highlighting the rapid growth of machine learning in climate modelling and model use for climate risk decisions.
Review of deep learning solar forecasting from sky cameras, satellites and weather stations that tracks cloud displacement, outlining gaps relative to NWP.
Applications & impactsUniversity of Cambridge (Aardvark)Stanford University
CNNs with partial dependence analysis reveal nonlinear soil moisture–temperature coupling across mid-latitude regions using reanalysis data.
Climate analysis & attributionStanford UniversityColorado State University (incl. CIRA)
Modified light-weight Context Guided CNN with weighted loss raises recall for tropical cyclone detection in climate simulation data.
ANNs trained on climate model output estimate observed timing of 1.5°C and 2°C warming, with a 2033–2035 central estimate for 1.5°C.
Climate analysis & attributionColorado State University (incl. CIRA)Stanford University
CNNs predict daily maximum temperature from soil moisture and geopotential, and partial dependence reveals nonlinear soil moisture-temperature coupling across mid-latitude regions.
Climate analysis & attributionStanford UniversityColorado State University (incl. CIRA)
Calibrate-emulate-sample with a Gaussian process emulator quantifies gravity wave parameterization uncertainty, showing QBO parametric uncertainty comparable to internal variability under 2xCO2.
Benchmarks off-the-shelf AutoML libraries on climate modelling and wind power forecasting; they fail to beat human-designed models, motivating spatiotemporal AutoML search spaces.
Calibrate-Emulate-Sample with a Gaussian process emulator quantifies gravity wave parameterization uncertainty; QBO parametric uncertainty is comparable to 2xCO2 internal variability.
Neural networks trained on SMAP soil moisture project Southeast Asian peat drying under RCP8.5, finding decreases about half as strong as recent El Niño droughts.
Applications & impactsMassachusetts Institute of Technology (MIT)NASAStanford University
Neural networks emulate a non-orographic gravity wave drag scheme and generalize to reproduce the QBO and the climate response to increased CO2 when coupled online.
Hybrid physics-ML & parameterizationNASAStanford UniversityNew York University (M2LInES)
A CNN identifies 91% of Midwest extreme precipitation events from circulation patterns and finds increasing frequency of the associated patterns over two decades.
Surveys physics-informed ML approaches with ten case studies for emulating, downscaling and forecasting weather and climate, reporting greater consistency and data efficiency.
Explainability, uncertainty & physicsCalifornia Institute of Technology (Caltech)Lawrence Berkeley National Laboratory / NERSCNVIDIAStanford UniversityUniversity of California, Berkeley
A cloudiness-classified CNN framework nowcasts PV output from sky images, reaching 2.20 kW RMSE and improving on an end-to-end CNN by 6%.
Explores deep learning input and output configurations for short-term solar power forecasts; the target is energy output rather than weather.
Position paper surveying how machine learning can help climate change mitigation and adaptation across sectors such as energy and disaster management; not focused on weather or climate science.
Applications & impactsETH ZurichGoogle DeepMindGoogle ResearchLawrence Berkeley National Laboratory / NERSCMassachusetts Institute of Technology (MIT)New York University (M2LInES)Stanford UniversityUniversity of California, Berkeley
Low-dimensional statistical emulator of the MAGICC climate model, built from designed CO2 scenarios, gives out-of-sample temperature errors below 2%.
Coupled CNN-RNN estimates local wind speed (0.75-11 m/s) from video of flags and trees, matching cup anemometer measurements.
Observations & remote sensingCalifornia Institute of Technology (Caltech)Stanford University
Opinion article arguing machine learning should be adopted for S2S forecasting, showing statistical methods beat dynamical models on a polar vortex case.
Sub-seasonal to seasonalMicrosoft Research (AI for Science)Massachusetts Institute of Technology (MIT)Potsdam Institute for Climate Impact Research (PIK)Stanford University
Ensemble of regression models with the SubseasonalRodeo dataset beats the top Rodeo entry and improves debiased CFSv2 precipitation skill by 129-169%.
Sub-seasonal to seasonalMicrosoft Research (AI for Science)Stanford University
Deep Gaussian process with CNN or LSTM features predicts county-level crop yields from remote sensing; the target is agriculture, not weather or climate.
