Neural networks for rainfall forecasting by atmospheric downscaling

J. Olsson, C. B. Uvo, K. Jinno, A. Kawamura, Koji Nishiyama, N. Koreeda, T. Nakashima, O. Morita

Research output: Contribution to journalArticle

60 Citations (Scopus)

Abstract

Several studies have used artificial neural networks (NNs) to estimate local or regional precipitation/rainfall on the basis of relationships with coarse-resolution atmospheric variables. None of these experiments satisfactorily reproduced temporal intermittency and variability in rainfall. We attempt to improve performance by using two approaches: (1) couple two NNs in series, the first to determine rainfall occurrence, and the second to determine rainfall intensity during rainy periods; and (2) categorize rainfall into intensity categories and train the NN to reproduce these rather than the actual intensities. The experiments focused on estimating 12-h mean rainfall in the Chikugo River basin, Kyushu Island, southern Japan, from large-scale values of wind speeds at 850 hPa and precipitable water. The results indicated that (1) two NNs in series may greatly improve the reproduction of intermittency; (2) longer data series are required to reproduce variability; (3) intensity categorization may be useful for probabilistic forecasting; and (4) overall performance in this region is better during winter and spring than during summer and autumn.

Original languageEnglish
Pages (from-to)1-12
Number of pages12
JournalJournal of Hydrologic Engineering
Volume9
Issue number1
DOIs
Publication statusPublished - Jan 2004

Fingerprint

downscaling
Rain
Neural networks
rainfall
precipitable water
precipitation intensity
Precipitation (meteorology)
artificial neural network
train
Catchments
river basin
experiment
wind velocity
autumn
Rivers
Experiments
winter
summer
Water

All Science Journal Classification (ASJC) codes

  • Environmental Chemistry
  • Civil and Structural Engineering
  • Water Science and Technology
  • Environmental Science(all)

Cite this

Neural networks for rainfall forecasting by atmospheric downscaling. / Olsson, J.; Uvo, C. B.; Jinno, K.; Kawamura, A.; Nishiyama, Koji; Koreeda, N.; Nakashima, T.; Morita, O.

In: Journal of Hydrologic Engineering, Vol. 9, No. 1, 01.2004, p. 1-12.

Research output: Contribution to journalArticle

Olsson, J, Uvo, CB, Jinno, K, Kawamura, A, Nishiyama, K, Koreeda, N, Nakashima, T & Morita, O 2004, 'Neural networks for rainfall forecasting by atmospheric downscaling', Journal of Hydrologic Engineering, vol. 9, no. 1, pp. 1-12. https://doi.org/10.1061/(ASCE)1084-0699(2004)9:1(1)
Olsson, J. ; Uvo, C. B. ; Jinno, K. ; Kawamura, A. ; Nishiyama, Koji ; Koreeda, N. ; Nakashima, T. ; Morita, O. / Neural networks for rainfall forecasting by atmospheric downscaling. In: Journal of Hydrologic Engineering. 2004 ; Vol. 9, No. 1. pp. 1-12.
@article{33409d54a2364d939b73b69abbd7e978,
title = "Neural networks for rainfall forecasting by atmospheric downscaling",
abstract = "Several studies have used artificial neural networks (NNs) to estimate local or regional precipitation/rainfall on the basis of relationships with coarse-resolution atmospheric variables. None of these experiments satisfactorily reproduced temporal intermittency and variability in rainfall. We attempt to improve performance by using two approaches: (1) couple two NNs in series, the first to determine rainfall occurrence, and the second to determine rainfall intensity during rainy periods; and (2) categorize rainfall into intensity categories and train the NN to reproduce these rather than the actual intensities. The experiments focused on estimating 12-h mean rainfall in the Chikugo River basin, Kyushu Island, southern Japan, from large-scale values of wind speeds at 850 hPa and precipitable water. The results indicated that (1) two NNs in series may greatly improve the reproduction of intermittency; (2) longer data series are required to reproduce variability; (3) intensity categorization may be useful for probabilistic forecasting; and (4) overall performance in this region is better during winter and spring than during summer and autumn.",
author = "J. Olsson and Uvo, {C. B.} and K. Jinno and A. Kawamura and Koji Nishiyama and N. Koreeda and T. Nakashima and O. Morita",
year = "2004",
month = "1",
doi = "10.1061/(ASCE)1084-0699(2004)9:1(1)",
language = "English",
volume = "9",
pages = "1--12",
journal = "Journal of Hydrologic Engineering - ASCE",
issn = "1084-0699",
publisher = "American Society of Civil Engineers (ASCE)",
number = "1",

}

TY - JOUR

T1 - Neural networks for rainfall forecasting by atmospheric downscaling

AU - Olsson, J.

AU - Uvo, C. B.

AU - Jinno, K.

AU - Kawamura, A.

AU - Nishiyama, Koji

AU - Koreeda, N.

AU - Nakashima, T.

AU - Morita, O.

PY - 2004/1

Y1 - 2004/1

N2 - Several studies have used artificial neural networks (NNs) to estimate local or regional precipitation/rainfall on the basis of relationships with coarse-resolution atmospheric variables. None of these experiments satisfactorily reproduced temporal intermittency and variability in rainfall. We attempt to improve performance by using two approaches: (1) couple two NNs in series, the first to determine rainfall occurrence, and the second to determine rainfall intensity during rainy periods; and (2) categorize rainfall into intensity categories and train the NN to reproduce these rather than the actual intensities. The experiments focused on estimating 12-h mean rainfall in the Chikugo River basin, Kyushu Island, southern Japan, from large-scale values of wind speeds at 850 hPa and precipitable water. The results indicated that (1) two NNs in series may greatly improve the reproduction of intermittency; (2) longer data series are required to reproduce variability; (3) intensity categorization may be useful for probabilistic forecasting; and (4) overall performance in this region is better during winter and spring than during summer and autumn.

AB - Several studies have used artificial neural networks (NNs) to estimate local or regional precipitation/rainfall on the basis of relationships with coarse-resolution atmospheric variables. None of these experiments satisfactorily reproduced temporal intermittency and variability in rainfall. We attempt to improve performance by using two approaches: (1) couple two NNs in series, the first to determine rainfall occurrence, and the second to determine rainfall intensity during rainy periods; and (2) categorize rainfall into intensity categories and train the NN to reproduce these rather than the actual intensities. The experiments focused on estimating 12-h mean rainfall in the Chikugo River basin, Kyushu Island, southern Japan, from large-scale values of wind speeds at 850 hPa and precipitable water. The results indicated that (1) two NNs in series may greatly improve the reproduction of intermittency; (2) longer data series are required to reproduce variability; (3) intensity categorization may be useful for probabilistic forecasting; and (4) overall performance in this region is better during winter and spring than during summer and autumn.

UR - http://www.scopus.com/inward/record.url?scp=1642578865&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=1642578865&partnerID=8YFLogxK

U2 - 10.1061/(ASCE)1084-0699(2004)9:1(1)

DO - 10.1061/(ASCE)1084-0699(2004)9:1(1)

M3 - Article

VL - 9

SP - 1

EP - 12

JO - Journal of Hydrologic Engineering - ASCE

JF - Journal of Hydrologic Engineering - ASCE

SN - 1084-0699

IS - 1

ER -