基于SARIMA模型的近岸海表溫度短期預(yù)報研究

作者：趙強  王擎宇  舒志光

單位：自然資源部寧波海洋中心, 浙江 寧波 315012

關(guān)鍵詞：周期性自回歸積分滑動平均方法 統(tǒng)計預(yù)報 海表溫度 預(yù)報

分類號：P731.31

出版年·卷·期（頁碼）：2024·41·第一期（42-49）

摘要：

基于石浦海洋站實測數(shù)據(jù)，采用周期性自回歸積分滑動平均方法（SARIMA）構(gòu)建了逐時海表溫度短期預(yù)報模型，根據(jù)觀測數(shù)據(jù)的周期特征和模型預(yù)報誤差比選確定了模型參數(shù)。結(jié)果表明：與采用逐時觀測數(shù)據(jù)作為輸入的模型相比，采用逐0.5 h內(nèi)插數(shù)據(jù)構(gòu)建的SARIMA模型的預(yù)報結(jié)果與實測數(shù)據(jù)間的相位更為一致，預(yù)報誤差更小，但進(jìn)一步將輸入數(shù)據(jù)的時間分辨率提高，72 h逐時預(yù)報精度提升不明顯；研究還發(fā)現(xiàn)模型預(yù)報誤差總體隨輸入數(shù)據(jù)時長的減小而增大；采用366 d逐0.5 h數(shù)據(jù)構(gòu)建的SARIMA (2,0,2)(2,1,0)25模型的預(yù)報結(jié)果較優(yōu)，0～24 h、24～48 h、48～72 h預(yù)報的平均絕對誤差分別為0.176℃、0.350℃、0.520℃，相應(yīng)的均方根誤差分別為0.217℃、0.396℃、0.567℃。

Based on the Sea Surface Temperature (SST) data of Shipu Station, time-series model of Seasonal AutoRegressive Integrated Moving Average (SARIMA) was used to construct a short-term forecasting model for hourly SST. Model parameters were determined according to the periodic of the data and the model forecasting errors. Compared to the model with original hourly input data, the model with interpolated half-hourly input data shows better performance, and the phases of the forecasts have a better consistent with the observations. Using higher temporal resolution of the input data shows no obvious improvement of the accuracy of the 72 h hourly SST forecasts. The results also show that the forecasting error increases with the reduction of the training data length. SARIMA(2, 0, 2) (2, 1, 0)25 model with 366-day interpolated half-hourly SST data shows the best forecasting accuracy. The mean absolute errors of 0～24 h, 24～48 h and 48～72 h forecasts are 0.176 ℃, 0.350 ℃ and 0.520 ℃, the corresponding root mean square error are 0.217 ℃, 0.396 ℃ and 0.567 ℃, respectively.

參考文獻(xiàn)：

[1] 張建華. 海溫預(yù)報知識講座: 第一講海水溫度預(yù)報概況[J]. 海洋預(yù)報, 2003, 20(4): 81-85. ZHANG J H. Lecture on sea water temperature prediction: lecture 1 overview of sea water temperature prediction[J]. Marine Forecasts, 2003, 20(4): 81-85. [2] 匡曉迪, 王兆毅, 張苗茵, 等. 基于BP神經(jīng)網(wǎng)絡(luò)方法的近岸數(shù)值海溫預(yù)報釋用技術(shù)[J]. 海洋與湖沼, 2016, 47(6): 1107-1115. KUANG X D, WANG Z Y, ZHANG M Y, et al. An interpretation scheme of numerical near-shore sea-water temperature forecast based on BPNN[J]. Oceanologia et Limnologia Sinica, 2016, 47(6): 1107-1115. [3] 韓玉康, 余丹丹, 申曉瑩, 等. HYCOM模式SST的預(yù)報誤差訂正[J]. 海洋預(yù)報, 2018, 35(3): 76-80. HAN Y K, YU D D, SHEN X Y, et al. Study on the correction of SST prediction of HYCOM[J]. Marine Forecasts, 2018, 35(3): 76-80. [4] 張建華. 海溫預(yù)報知識講座: 第二講數(shù)理統(tǒng)計方法在海溫預(yù)報中的應(yīng)用[J]. 海洋預(yù)報, 2004, 21(1): 85-90. ZHANG J H. Lecture on sea water temperature prediction: lecture 2 application of mathematical statistical methods on sea water temperature prediction[J]. Marine Forecasts, 2004, 21(1): 85-90. [5] BOX G E P, JENKINS G M, REINSEL G C. Time series analysis: forecasting and control[M]. 3rd ed. Englewood Cliffs, NJ: Prentice Hall, 1994. [6] LIU X L, LIN Z, FENG Z M. Short-term offshore wind speed forecast by seasonal ARIMA—A comparison against GRU and LSTM[J]. Energy, 2021, 227: 120492. [7] 俞肇元, 袁林旺, 謝志仁, 等. 基于SSA和AR模型的海面變化預(yù)測試驗[J]. 海洋湖沼通報, 2007(4): 14-20. YU Z Y, YUAN L W, XIE Z R, et al. Prediction experiment of sealevel change based on the SSA and AR model[J]. Transactions of Oceanology and Limnology, 2007(4): 14-20. [8] HALMOVA D, PEKAROVA P, PEKAR J, et al. The simulation of the water temperature rising using ARIMA models[J]. WSEAS Transactions on Heat and Mass Transfer, 2016, 11: 46-55. [9] 張瑩, 譚艷春, 彭發(fā)定, 等. 基于EEMD和ARIMA的海溫預(yù)測模型研究[J]. 海洋學(xué)研究, 2019, 37(1): 9-14. ZHANG Y, TAN Y C, PENG F D, et al. Study on time series prediction model of sea surface temperature based on Ensemble Empirical Mode Decomposition and Autoregressive Integrated Moving Average[J]. Journal of Marine Sciences, 2019, 37(1): 9-14. [10] 徐麗麗, 余駿, 高鑫鑫, 等. 一種基于時間序列分析的赤潮預(yù)測新方法研究——以浙江海域為例[J]. 海洋預(yù)報, 2020, 37(5): 95-103. XU L L, YU J, GAO X X, et al. Study on the red tide prediction based on time series analysis—A case study in Zhejiang sea area [J]. Marine Forecasts, 2020, 37(5): 95-103. [11] 張振全, 李醒飛, 楊少波. 基于AR-SVR模型的有效波高短期預(yù)測[J]. 太陽能學(xué)報, 2021, 42(7): 15-20. ZHANG Z Q, LI X F, YANG S B. Short-term prediction of significant wave height based on AR-SVR model[J]. Acta Energiae Solaris Sinica, 2021, 42(7): 15-20. [12] 劉嬌, 史國友, 朱凱歌, 等. 基于調(diào)和分析和ARIMA-SVR的組合潮汐預(yù)測模型[J]. 上海海事大學(xué)學(xué)報, 2019, 40(3): 93-99. LIU J, SHI G Y, ZHU K G, et al. A combined tide prediction model based on harmonic analysis and ARIMA-SVR[J]. Journal of Shanghai Maritime University, 2019, 40(3): 93-99. [13] ARUNKUMAR K E, KALAGA D V, SAI KUMAR C M, et al. Forecasting the dynamics of cumulative COVID-19 cases (confirmed, recovered and deaths) for top-16 countries using statistical machine learning models: auto-Regressive Integrated Moving Average (ARIMA) and Seasonal Auto-Regressive Integrated Moving Average (SARIMA) [J]. Applied Soft Computing, 2021, 103: 107161. [14] 耿宏, 王偉, 邢承濱. 基于ARMA模型的潮位擬合及內(nèi)插研究[J]. 海洋測繪, 2021, 41(5): 17-20. GENG H, WANG W, XING C B. Research on tide fitting and interpolation based on ARMA model[J]. Hydrographic Surveying and Charting, 2021, 41(5): 17-20. [15] DICKEY D A, FULLER W A. Likelihood ratio statistics for autoregressive time series with a unit root[J]. Econometrica, 1981, 49(4): 1057-1072. [16] 白彬人, 宋家喜. 中國近海沿岸海溫多時間尺度變率及影響其變化的天氣氣候因素[J]. 海洋預(yù)報, 2005, 22(4): 78-88. BAI B R, SONG J X. Multi-temporal sea temperature variability in the coastal areas of China and their affecting atmospheric and climatic factors[J]. Marine Forecasts, 2005, 22(4): 78-88. [17] 國家市場監(jiān)督管理總局, 國家標(biāo)準(zhǔn)化管理委員會. GB/T 41165-2021海洋預(yù)報結(jié)果準(zhǔn)確性檢驗評估方法[S]. 北京: 中國標(biāo)準(zhǔn)出版社, 2021. State Administration for Market Regulation, Standardization Administration. GB/T 41165-2021 Accuracy evaluation methods of marine forecast results[S]. Beijing: Standards Press of China, 2021. [18] 李燕, 張建華, 劉欽政, 等. 單站海溫短期預(yù)報自動化[J]. 海洋預(yù)報, 2007, 24(4): 33-41. LI Y, ZHANG J H, LIU Q Z, et al. The automation of single sea station's surface sea temperature short term forecasting[J]. Marine Forecasts, 2007, 24(4): 33-41. [19] 林小剛, 王兆毅, 李競時, 等. 基于LSTM神經(jīng)網(wǎng)絡(luò)方法的粵東近岸海溫預(yù)報[J]. 海洋預(yù)報, 2022, 39(5): 27-36. LIN X G, WANG Z Y, LI J S, et al. Sea temperature forecasting based on LSTM neural network along the coast of eastern Guangdong[J]. Marine Forecasts, 2022, 39(5): 27-36. [20] 王兆毅, 李云, 王旭. 中國近岸海域基礎(chǔ)預(yù)報單元海溫預(yù)報指導(dǎo)產(chǎn)品研制[J]. 海洋預(yù)報, 2020, 37(4): 59-65. WANG Z Y, LI Y, WANG X. Development of forecast guidance product for sea temperature of basic forecast units in the Chinese coastal waters[J]. Marine Forecasts, 2020, 37(4): 59-65.

服務(wù)與反饋：

【文章下載】【發(fā)表評論】【查看評論】【加入收藏】