Abnormal Water Temperature Prediction Model Near the Korean Peninsula Using LSTM SCOPUS KCI

DC Field Value Language
dc.contributor.author Choi, Hey Min -
dc.contributor.author Kim, Min-Kyu -
dc.contributor.author Yang, Hyun -
dc.date.accessioned 2022-07-04T00:50:01Z -
dc.date.available 2022-07-04T00:50:01Z -
dc.date.created 2022-07-04 -
dc.date.issued 2022-06 -
dc.identifier.issn 1225-6161 -
dc.identifier.uri https://sciwatch.kiost.ac.kr/handle/2020.kiost/43049 -
dc.description.abstract Sea surface temperature (SST) is a factor that greatly influences ocean circulation and ecosystems in the Earth system. As global warming causes changes in the SST near the Korean Peninsula, abnormal water temperature phenomena (high water temperature, low water temperature) occurs, causing continuous damage to the marine ecosystem and the fishery industry. Therefore, this study proposes a methodology to predict the SST near the Korean Peninsula and prevent damage by predicting abnormal water temperature phenomena. The study area was set near the Korean Peninsula, and ERA5 data from the European Center for Medium-Range Weather Forecasts (ECMWF) was used to utilize SST data at the same time period. As a research method, Long Short-Term Memory (LSTM) algorithm specialized for time series data prediction among deep learning models was used in consideration of the time series characteristics of SST data. The prediction model predicts the SST near the Korean Peninsula after 1- to 7-days and predicts the high water temperature or low water temperature phenomenon. To evaluate the accuracy of SST prediction, Coefficient of determination (R2), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE) indicators were used. The summer (JAS) 1-day prediction result of the prediction model, R2=0.996, RMSE=0.119°C, MAPE=0.352% and the winter (JFM) 1-day prediction result is R2=0.999, RMSE=0.063°C, MAPE=0.646%. Using the predicted SST, the accuracy of abnormal sea surface temperature prediction was evaluated with an F1 Score (F1 Score=0.98 for high water temperature prediction in summer (2021/08/05), F1 Score=1.0 for low water temperature prediction in winter (2021/02/19)). As the prediction period increased, the prediction model showed a tendency to underestimate the SST, which also reduced the accuracy of the abnormal water temperature prediction. Therefore, it is judged that it is necessary to analyze the cause of underestimation of the predictive model in the future and study to improve the prediction accuracy. 해수면 온도(Sea surface temperature, SST)는 지구시스템에서 해양의 순환과 생태계에 큰 영향을 주는 요소이다. 지구온난화로 한반도 근해 해수면 온도에 변화가 생기면서 이상 수온(고수온, 저수온) 현상이 발생하여 해양생태계와 수산업 피해를 지속적으로 발생시키고 있다. 따라서 본 연구는 한반도 근해 해수면 온도를 예측하여 이상 수온 현상 예측으로 피해를 예방하는 방법론을 제안한다. 연구 지역은 한반도 근해로 설정하여 동시간대 해수면 온도 데이터를 사용하기 위해 Europe Centre for Medium-Range Weather Forecasts (ECMWF)의ERA5 자료를 사용하였다. 연구방법으로는 해수면 온도 데이터의 시계열 특징을 고려하여 딥러닝 모델 중 시계열 데이터 예측에 특화된 Long Short-Term Memory (LSTM) 알고리즘을 이용하였다. 예측 모델은 1~7일 이후한반도 근해 해수면 온도를 예측하고 고수온(High water temperature, HWT) 혹은 저수온(Low water temperature,LWT) 현상을 예측한다. 해수면 온도 예측 정확도 평가를 위해 결정계수(Coefficient of determination, R2), 평균제곱근 편차(Root Mean Squared Error, RMSE), 평균 절대 백분율 오차(Mean Absolute Percentage Error, MAPE)지표를 사용하였다. 예측 모델의 여름철(JAS) 1일 예측 결과는 R2=0.996, RMSE=0.119°C, MAPE=0.352% 이고,겨울철(JFM) 1일 예측 결과는 R2=0.999, RMSE=0.063°C, MAPE=0.646% 이었다. 예측한 해수면 온도를 이용하여 이상 수온 예측 정확도 평가를 F1 Score로 수행하였다(여름철(2021/08/05) 고수온 예측 결과 F1 Score=0.98,겨울철(2021/02/19) 저수온 예측 결과 F1 Score=1.0). 예측 기간이 증가하면서 예측 모델이 해수면 온도를 과소추정하는 경향을 보여주었고, 이로 인해 이상 수온 예측 정확도 또한 낮아졌다. 따라서, 향후 예측 모델의 과소추정 원인을 분석하고 예측 정확도 향상을 위한 연구가 필요할 것으로 판단된다. -
dc.description.uri 3 -
dc.language Korean -
dc.publisher 대한원격탐사학회 -
dc.title Abnormal Water Temperature Prediction Model Near the Korean Peninsula Using LSTM -
dc.title.alternative LSTM을 이용한 한반도 근해 이상수온 예측모델 -
dc.type Article -
dc.citation.endPage 282 -
dc.citation.startPage 265 -
dc.citation.title Korean Journal of Remote Sensing -
dc.citation.volume 38 -
dc.citation.number 3 -
dc.contributor.alternativeName 최혜민 -
dc.identifier.bibliographicCitation Korean Journal of Remote Sensing, v.38, no.3, pp.265 - 282 -
dc.identifier.doi 10.7780/kjrs.2022.38.3.4 -
dc.identifier.scopusid 2-s2.0-85134669250 -
dc.identifier.kciid ART002849904 -
dc.description.journalClass 3 -
dc.description.isOpenAccess N -
dc.subject.keywordAuthor Sea surface temperature -
dc.subject.keywordAuthor High water temperature -
dc.subject.keywordAuthor Low water temperature -
dc.subject.keywordAuthor Long short-term memory -
dc.description.journalRegisteredClass scopus -
dc.description.journalRegisteredClass kci -
Appears in Collections:
Marine Digital Resources Department > Korea Ocean Satellite Center > 1. Journal Articles
Files in This Item:
There are no files associated with this item.

qrcode

Items in ScienceWatch@KIOST are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse