Analysis on the Occurrence of Tropical Cyclone in the South Pacific Region Using Recurrent Neural Network with LSTM

Weather prediction over the years has been a challenge for the meteorological centers in the South Pacific region. This paper presents Recurrent Neural Network (RNN) Architecture with Long Short Term Memory (LSTM) times-series weather data for prediction. From the gathered dataset, the Sea Surface Temperature (SST) is studied since it is known to be the foundation of the cyclone formation. This paper focuses on two scenarios. The first part is predicting upcoming SST using dataset from January 2013 to December 2017. The second part is taking out data of two different cyclones and predicting the SST for the next 14 days. Once the SST prediction is made, the predicted SST is compared with SST in the dataset for those 14 days. The main aim of this paper is to predict the SST using RNN and LSTM to anticipate the occurrence of tropical cyclones. The paper will focus on the reason for this study, a discussion of the model used, how the cyclones are formed, regarding the current threshold, the analysis of the dataset and lastly, the results from the experiment carried out.

[1]  Sepp Hochreiter,et al.  The Vanishing Gradient Problem During Learning Recurrent Neural Nets and Problem Solutions , 1998, Int. J. Uncertain. Fuzziness Knowl. Based Syst..

[2]  Chaker El Amrani,et al.  Sequence to Sequence Weather Forecasting with Long Short-Term Memory Recurrent Neural Networks , 2016 .

[3]  J. Chan,et al.  The Effects of the Full Coriolis Force on the Structure and Motion of a Tropical Cyclone. Part I: Effects due to Vertical Motion , 2005 .

[4]  D. Raymond,et al.  Thermodynamics of tropical cyclogenesis in the northwest Pacific , 2011 .

[5]  Makarand Deo,et al.  Prediction of Sea Surface Temperature by Combining Numerical and Neural Techniques , 2016 .

[6]  C. Ebinger,et al.  Comparison of dike intrusions in an incipient seafloor‐spreading segment in Afar, Ethiopia: Seismicity perspectives , 2011 .

[7]  P. Dash,et al.  Towards Dependence of Tropical Cyclone Intensity on Sea Surface Temperature and Its Response in a Warming World , 2016 .

[8]  Sea‐surface temperatures and tropical cyclones in the Atlantic basin , 2006 .

[9]  Isaac Luginaah,et al.  Perceived Self-Efficacy and Adaptation to Climate Change in Coastal Cambodia , 2015 .

[10]  M. Montgomery.  Recent Advances in Tropical Cyclogenesis , 2016 .

[11]  Junyu Dong,et al.  Prediction of Sea Surface Temperature Using Long Short-Term Memory , 2017, IEEE Geoscience and Remote Sensing Letters.

[12]  S. Aparna,et al.  Prediction of daily sea surface temperature using artificial neural networks , 2018 .

[13]  Zachary Chase Lipton.  A Critical Review of Recurrent Neural Networks for Sequence Learning , 2015, ArXiv.

[14]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[15]  R. Dare,et al.  Sea Surface Temperature Thresholds for Tropical Cyclone Formation , 2015 .

[16]  J. McBride,et al.  The Threshold Sea Surface Temperature Condition for Tropical Cyclogenesis , 2011 .

[17]  W. M. Gray,et al.  GLOBAL VIEW OF THE ORIGIN OF TROPICAL DISTURBANCES AND STORMS , 1968 .

[18]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[19]  A. Hedin,et al.  Atomic nitrogen densities in the thermosphere , 1976 .