An Arctic Evaluation of the Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) Release 1 Neural Network Cloud Mask

Abstract The Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) operational Release 1 (R01) neural network cloud mask (2B‐MSK) is described. We use Atmospheric Infrared Sounder (AIRS), Cross‐track Infrared Microwave Sounder Suite (CrIMSS), Visible Infrared Imaging Radiometer Suite (VIIRS), and CloudSat/CALIPSO cloud products to evaluate 2B‐MSK for Arctic data. Climatological comparisons against yearly CloudSat/CALIPSO climatologies suggest that PREFIRE is more sensitive to cloud occurrence than estimated from pre‐launch simulations. Average footprint‐scale cloud detection hit rates ( H ) are nearly identical for SAT1 and SAT2, ranging between 72% and 75% against AIRS and CrIMSS and 80%–82% against VIIRS. Higher (lower) H is found over ocean and land (snow and ice) surfaces. A decrease of H by a few percent is found for the largest temporal and spatial offsets between coincidences, while a decrease of 5%–8% is found for the highest AIRS, CrIMSS, and VIIRS satellite viewing angles. Thus, H is highest for the tightest subset of temporal and spatial collocations closest to nadir. A prototype ensemble‐based version of 2B‐MSK more fully captures uncertainties arising from limited sampling in R01 training sets. Ensemble mean values lead to higher H , particularly over snow and ice surfaces, and reduced cross‐track discrepancies in PREFIRE‐detected clouds. Consistent with previous climatological studies, PREFIRE observes larger cloud amounts with stronger surface winds, increased vertical wind shear, and ascending air. This study documents the first known cloud climatology derived from spaceborne measurements of emission spectra that include far‐infrared wavelengths longer than 15 μm.

Authors

Institutions

Publication Details

Journal
Journal of Geophysical Research Atmospheres
Published
2026-11-14
DOI
https://doi.org/10.1029/2026jd047208
Primary Topic
Atmospheric aerosols and clouds
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

An Arctic Evaluation of the Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) Release 1 Neural Network Cloud Mask

Aronne J. Merrelli, Kyle S. Mattingly, Nathaniel B. Miller, Brian Hannon Kahn et al.
2 links
Journal of Geophysical Research Atmospheres
Atmospheric aerosols and clouds
article

An Arctic Evaluation of the Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) Release 1 Neural Network Cloud Mask

Aronne J. Merrelli, Kyle S. Mattingly, Nathaniel B. Miller, Brian Hannon Kahn, Xianglei Huang, Tristan L’Ecuyer, Timothy I. Michaels, Brian J. Drouin, Qing Yue, Cameron Bertossa, Erin Wagner Hokanson, Xiuhong Chen, Ben Konkel
article en

Abstract

Abstract The Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) operational Release 1 (R01) neural network cloud mask (2B‐MSK) is described. We use Atmospheric Infrared Sounder (AIRS), Cross‐track Infrared Microwave Sounder Suite (CrIMSS), Visible Infrared Imaging Radiometer Suite (VIIRS), and CloudSat/CALIPSO cloud products to evaluate 2B‐MSK for Arctic data. Climatological comparisons against yearly CloudSat/CALIPSO climatologies suggest that PREFIRE is more sensitive to cloud occurrence than estimated from pre‐launch simulations. Average footprint‐scale cloud detection hit rates ( H ) are nearly identical for SAT1 and SAT2, ranging between 72% and 75% against AIRS and CrIMSS and 80%–82% against VIIRS. Higher (lower) H is found over ocean and land (snow and ice) surfaces. A decrease of H by a few percent is found for the largest temporal and spatial offsets between coincidences, while a decrease of 5%–8% is found for the highest AIRS, CrIMSS, and VIIRS satellite viewing angles. Thus, H is highest for the tightest subset of temporal and spatial collocations closest to nadir. A prototype ensemble‐based version of 2B‐MSK more fully captures uncertainties arising from limited sampling in R01 training sets. Ensemble mean values lead to higher H , particularly over snow and ice surfaces, and reduced cross‐track discrepancies in PREFIRE‐detected clouds. Consistent with previous climatological studies, PREFIRE observes larger cloud amounts with stronger surface winds, increased vertical wind shear, and ascending air. This study documents the first known cloud climatology derived from spaceborne measurements of emission spectra that include far‐infrared wavelengths longer than 15 μm.

Journal of Geophysical Research AtmospheresVol. 131(22)
Jet Propulsion Laboratory (US), University of Wisconsin–Madison (US), University of Michigan (US), Space Science and Engineering Center (US), Cooperative Institute for Meteorological Satellite Studies (US)
Climate action
Atmospheric aerosols and clouds
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon