Prediction of daily sea surface temperature using artificial neural networks

S. G. Aparna, Selrina D’souza, N. B. Arjun

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

We present an artificial neural network model to predict the sea surface temperature (SST) and delineate SST fronts in the northeastern Arabian Sea. The predictions are made one day in advance, using current day’s SST for predicting the SST of the next day. The model is used to predict the SST map for every single day during 2013–2015. The results show that more than 75% of the time the model error is ≤ ±0.5ºC. For the years 2014 and 2015, 80% of the predictions had an error ≤±0.5ºC. The model performance is dependent on the availability of data during the previous days. Thus during the summer monsoon months, when the data availability is comparatively less, the errors in the prediction are slightly higher. The model is also able to capture SST fronts.

Original languageEnglish
Pages (from-to)4214-4231
Number of pages18
JournalInternational Journal of Remote Sensing
Volume39
Issue number12
DOIs
Publication statusPublished - 18-06-2018
Externally publishedYes

Fingerprint

artificial neural network
sea surface temperature
prediction
monsoon
summer

All Science Journal Classification (ASJC) codes

  • Earth and Planetary Sciences(all)

Cite this

Aparna, S. G. ; D’souza, Selrina ; Arjun, N. B. / Prediction of daily sea surface temperature using artificial neural networks. In: International Journal of Remote Sensing. 2018 ; Vol. 39, No. 12. pp. 4214-4231.
@article{788c639ef72d4ef1a9d35fd498794799,
title = "Prediction of daily sea surface temperature using artificial neural networks",
abstract = "We present an artificial neural network model to predict the sea surface temperature (SST) and delineate SST fronts in the northeastern Arabian Sea. The predictions are made one day in advance, using current day’s SST for predicting the SST of the next day. The model is used to predict the SST map for every single day during 2013–2015. The results show that more than 75{\%} of the time the model error is ≤ ±0.5ºC. For the years 2014 and 2015, 80{\%} of the predictions had an error ≤±0.5ºC. The model performance is dependent on the availability of data during the previous days. Thus during the summer monsoon months, when the data availability is comparatively less, the errors in the prediction are slightly higher. The model is also able to capture SST fronts.",
author = "Aparna, {S. G.} and Selrina D’souza and Arjun, {N. B.}",
year = "2018",
month = "6",
day = "18",
doi = "10.1080/01431161.2018.1454623",
language = "English",
volume = "39",
pages = "4214--4231",
journal = "International Journal of Remote Sensing",
issn = "0143-1161",
publisher = "Taylor and Francis Ltd.",
number = "12",

}

Prediction of daily sea surface temperature using artificial neural networks. / Aparna, S. G.; D’souza, Selrina; Arjun, N. B.

In: International Journal of Remote Sensing, Vol. 39, No. 12, 18.06.2018, p. 4214-4231.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Prediction of daily sea surface temperature using artificial neural networks

AU - Aparna, S. G.

AU - D’souza, Selrina

AU - Arjun, N. B.

PY - 2018/6/18

Y1 - 2018/6/18

N2 - We present an artificial neural network model to predict the sea surface temperature (SST) and delineate SST fronts in the northeastern Arabian Sea. The predictions are made one day in advance, using current day’s SST for predicting the SST of the next day. The model is used to predict the SST map for every single day during 2013–2015. The results show that more than 75% of the time the model error is ≤ ±0.5ºC. For the years 2014 and 2015, 80% of the predictions had an error ≤±0.5ºC. The model performance is dependent on the availability of data during the previous days. Thus during the summer monsoon months, when the data availability is comparatively less, the errors in the prediction are slightly higher. The model is also able to capture SST fronts.

AB - We present an artificial neural network model to predict the sea surface temperature (SST) and delineate SST fronts in the northeastern Arabian Sea. The predictions are made one day in advance, using current day’s SST for predicting the SST of the next day. The model is used to predict the SST map for every single day during 2013–2015. The results show that more than 75% of the time the model error is ≤ ±0.5ºC. For the years 2014 and 2015, 80% of the predictions had an error ≤±0.5ºC. The model performance is dependent on the availability of data during the previous days. Thus during the summer monsoon months, when the data availability is comparatively less, the errors in the prediction are slightly higher. The model is also able to capture SST fronts.

UR - http://www.scopus.com/inward/record.url?scp=85059069589&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059069589&partnerID=8YFLogxK

U2 - 10.1080/01431161.2018.1454623

DO - 10.1080/01431161.2018.1454623

M3 - Article

VL - 39

SP - 4214

EP - 4231

JO - International Journal of Remote Sensing

JF - International Journal of Remote Sensing

SN - 0143-1161

IS - 12

ER -