TIME SERIES FORECASTING FOR AVERAGE TEMPERATURE IN 96041 STATION USING LONG SHORT-TERM MEMORY MODEL

  • Nancy Lusiana Damanik Indonesia Meteorology, Climatology and Geophysics Agency
  • Elida Pane Indonesia Meteorology, Climatology and Geophysics Agency
  • Kartika Dewi Indonesia Meteorology, Climatology and Geophysics Agency
  • Efrianses F. H. Sinaga Indonesia Meteorology, Climatology and Geophysics Agency
  • Jamaluddin Jamaluddin Universitas Methodist Indonesia http://orcid.org/0000-0001-7458-7885
  • Hiras Sinaga Indonesia Meteorology, Climatology and Geophysics Agency
  • Marzuki Sinambela Indonesia Meteorology, Climatology and Geophysics Agency
Keywords: Average Temperature, Forecasting, LSTM, Time-Series, 96041 Station

Abstract

An understanding of trends and prediction of average temperature as one of parameter weather and climate data for better water resource management and planning in a basin is very important. Explore weather trends using normal and local yearly average temperatures, compare and make observations. In this study, we will analyze local and normal average temperature data in 96041 Station based on observation station in situ. The main goal of this study to show the performance of the average temperature in a single station and to predict the average temperature data using the Long Short-Term Memory Model approach. Based on the result of normal data science of exploring temperature with local temperature correlation, we got the display of training curve, residual plot, and the scatter plot is shown using these codes. The good performance of 96041 Station had an MSE value of 0.01 and R2 value 0.98, close to zero will represent better quality of the predictor.

Published
2021-04-06

Most read articles by the same author(s)