AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

ClimART: A Benchmark Dataset for Emulating Atmospheric Radiative Transfer in Weather and Climate Models

Salva Rühling Cachay, Venkatesh Ramesh, Jason N. S. Cole, Howard Barker, David Rolnick

NeurIPS 2021 · Probabilistic and scientific modeling › Molecules and proteins · approach: Machine learning for physical simulation · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use 1D CNN instead of MLPMLP · ClimART pristine-sky, vertical avg.RMSE: 0.701 -> 0.303Improved
use GraphNet instead of MLPMLP · ClimART pristine-sky, vertical avg.RMSE: 0.701 -> 0.648Improved
use GraphNet instead of MLP for biasMLP · ClimART pristine-sky, vertical avg.MBE: -0.160 -> -0.001Improved
use GCN with line-graph instead of MLPMLP · ClimART pristine-sky, vertical avg.RMSE: 0.701 -> 1.209Hurt
learn the adjacency matrix (L-GCN) instead of fixed line graphGCN · ClimART pristine-sky, vertical avg.RMSE: 1.209 -> 0.878Improved
learn the adjacency matrix (L-GCN) instead of fixed line graphGCN · ClimART pristine-sky, TOARMSE: 0.815 -> 0.440Improved
run models on GPU instead of 4 CPUsMLP · ClimART speedinference time (s): 0.1643 -> 0.0016Improved
run models on GPU instead of 4 CPUsGraphNet · ClimART speedinference time (s): 28.1253 -> 0.1659Improved
use MLP on GPU instead of physics-based RT (64 CPUs)Physics-RT · ClimART speedinference time (s): 1.9817 -> 0.0016Improved
increase number of CPUs for physics RT from 2 to 64Physics-RT · ClimART speedinference time (s): 3.3919 -> 1.9817Improved
increase batch size from 512 to 8192ML models · ClimART speedinference throughput: 3.5xImproved
test on future climate conditions instead of present-dayMLP · ClimART OODRMSE: 1 -> 100Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy