New neural network cloud mask algorithm based on radiative transfer simulations

Nan Chen, Wei Li, Charles Gatebe, Tomonori Tanikawa, Masahiro Hori, Rigen Shimada, Teruo Aoki, Knut Stamnes

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Cloud detection and screening constitute critically important first steps required to derive many satellite data products. Traditional threshold-based cloud mask algorithms require a complicated design process and fine tuning for each sensor, and they have difficulties over areas partially covered with snow/ice. Exploiting advances in machine learning techniques and radiative transfer modeling of coupled environmental systems, we have developed a new, threshold-free cloud mask algorithm based on a neural network classifier driven by extensive radiative transfer simulations. Statistical validation results obtained by using collocated CALIOP and MODIS data show that its performance is consistent over different ecosystems and significantly better than the MODIS Cloud Mask (MOD35 C6) during the winter seasons over snow-covered areas in the mid-latitudes. Simulations using a reduced number of satellite channels also show satisfactory results, indicating its flexibility to be configured for different sensors. Compared to threshold-based methods and previous machine-learning approaches, this new cloud mask (i) does not rely on thresholds, (ii) needs fewer satellite channels, (iii) has superior performance during winter seasons in mid-latitude areas, and (iv) can easily be applied to different sensors.

Original languageEnglish
Pages (from-to)62-71
Number of pages10
JournalRemote Sensing of Environment
Volume219
DOIs
Publication statusPublished - Dec 15 2018

Fingerprint

Radiative transfer
neural networks
sensors (equipment)
radiative transfer
Masks
artificial intelligence
moderate resolution imaging spectroradiometer
Neural networks
snow
Satellites
Snow
simulation
Learning systems
Sensors
sensor
winter
MODIS
remote sensing
ice
Ecosystems

Keywords

  • Cloud mask
  • Machine learning
  • Radiative transfer

ASJC Scopus subject areas

  • Soil Science
  • Geology
  • Computers in Earth Sciences

Cite this

Chen, N., Li, W., Gatebe, C., Tanikawa, T., Hori, M., Shimada, R., ... Stamnes, K. (2018). New neural network cloud mask algorithm based on radiative transfer simulations. Remote Sensing of Environment, 219, 62-71. https://doi.org/10.1016/j.rse.2018.09.029

New neural network cloud mask algorithm based on radiative transfer simulations. / Chen, Nan; Li, Wei; Gatebe, Charles; Tanikawa, Tomonori; Hori, Masahiro; Shimada, Rigen; Aoki, Teruo; Stamnes, Knut.

In: Remote Sensing of Environment, Vol. 219, 15.12.2018, p. 62-71.

Research output: Contribution to journalArticle

Chen, N, Li, W, Gatebe, C, Tanikawa, T, Hori, M, Shimada, R, Aoki, T & Stamnes, K 2018, 'New neural network cloud mask algorithm based on radiative transfer simulations', Remote Sensing of Environment, vol. 219, pp. 62-71. https://doi.org/10.1016/j.rse.2018.09.029
Chen, Nan ; Li, Wei ; Gatebe, Charles ; Tanikawa, Tomonori ; Hori, Masahiro ; Shimada, Rigen ; Aoki, Teruo ; Stamnes, Knut. / New neural network cloud mask algorithm based on radiative transfer simulations. In: Remote Sensing of Environment. 2018 ; Vol. 219. pp. 62-71.
@article{994d7fedea634777ac48a24353e624cc,
title = "New neural network cloud mask algorithm based on radiative transfer simulations",
abstract = "Cloud detection and screening constitute critically important first steps required to derive many satellite data products. Traditional threshold-based cloud mask algorithms require a complicated design process and fine tuning for each sensor, and they have difficulties over areas partially covered with snow/ice. Exploiting advances in machine learning techniques and radiative transfer modeling of coupled environmental systems, we have developed a new, threshold-free cloud mask algorithm based on a neural network classifier driven by extensive radiative transfer simulations. Statistical validation results obtained by using collocated CALIOP and MODIS data show that its performance is consistent over different ecosystems and significantly better than the MODIS Cloud Mask (MOD35 C6) during the winter seasons over snow-covered areas in the mid-latitudes. Simulations using a reduced number of satellite channels also show satisfactory results, indicating its flexibility to be configured for different sensors. Compared to threshold-based methods and previous machine-learning approaches, this new cloud mask (i) does not rely on thresholds, (ii) needs fewer satellite channels, (iii) has superior performance during winter seasons in mid-latitude areas, and (iv) can easily be applied to different sensors.",
keywords = "Cloud mask, Machine learning, Radiative transfer",
author = "Nan Chen and Wei Li and Charles Gatebe and Tomonori Tanikawa and Masahiro Hori and Rigen Shimada and Teruo Aoki and Knut Stamnes",
year = "2018",
month = "12",
day = "15",
doi = "10.1016/j.rse.2018.09.029",
language = "English",
volume = "219",
pages = "62--71",
journal = "Remote Sensing of Environment",
issn = "0034-4257",
publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - New neural network cloud mask algorithm based on radiative transfer simulations

AU - Chen, Nan

AU - Li, Wei

AU - Gatebe, Charles

AU - Tanikawa, Tomonori

AU - Hori, Masahiro

AU - Shimada, Rigen

AU - Aoki, Teruo

AU - Stamnes, Knut

PY - 2018/12/15

Y1 - 2018/12/15

N2 - Cloud detection and screening constitute critically important first steps required to derive many satellite data products. Traditional threshold-based cloud mask algorithms require a complicated design process and fine tuning for each sensor, and they have difficulties over areas partially covered with snow/ice. Exploiting advances in machine learning techniques and radiative transfer modeling of coupled environmental systems, we have developed a new, threshold-free cloud mask algorithm based on a neural network classifier driven by extensive radiative transfer simulations. Statistical validation results obtained by using collocated CALIOP and MODIS data show that its performance is consistent over different ecosystems and significantly better than the MODIS Cloud Mask (MOD35 C6) during the winter seasons over snow-covered areas in the mid-latitudes. Simulations using a reduced number of satellite channels also show satisfactory results, indicating its flexibility to be configured for different sensors. Compared to threshold-based methods and previous machine-learning approaches, this new cloud mask (i) does not rely on thresholds, (ii) needs fewer satellite channels, (iii) has superior performance during winter seasons in mid-latitude areas, and (iv) can easily be applied to different sensors.

AB - Cloud detection and screening constitute critically important first steps required to derive many satellite data products. Traditional threshold-based cloud mask algorithms require a complicated design process and fine tuning for each sensor, and they have difficulties over areas partially covered with snow/ice. Exploiting advances in machine learning techniques and radiative transfer modeling of coupled environmental systems, we have developed a new, threshold-free cloud mask algorithm based on a neural network classifier driven by extensive radiative transfer simulations. Statistical validation results obtained by using collocated CALIOP and MODIS data show that its performance is consistent over different ecosystems and significantly better than the MODIS Cloud Mask (MOD35 C6) during the winter seasons over snow-covered areas in the mid-latitudes. Simulations using a reduced number of satellite channels also show satisfactory results, indicating its flexibility to be configured for different sensors. Compared to threshold-based methods and previous machine-learning approaches, this new cloud mask (i) does not rely on thresholds, (ii) needs fewer satellite channels, (iii) has superior performance during winter seasons in mid-latitude areas, and (iv) can easily be applied to different sensors.

KW - Cloud mask

KW - Machine learning

KW - Radiative transfer

UR - http://www.scopus.com/inward/record.url?scp=85054464404&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054464404&partnerID=8YFLogxK

U2 - 10.1016/j.rse.2018.09.029

DO - 10.1016/j.rse.2018.09.029

M3 - Article

AN - SCOPUS:85054464404

VL - 219

SP - 62

EP - 71

JO - Remote Sensing of Environment

JF - Remote Sensing of Environment

SN - 0034-4257

ER -