Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany
Local temperature forecasts for horizons up to 24 h are required in many applications. A common method to generate such forecasts is the Seasonal Autoregressive Integrated Moving Average (SARIMA) model or, much simpler, the naïve forecast. In this paper, we test whether deep neural networks are able...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2020-12-01
|
| Series: | Machine Learning with Applications |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666827020300074 |
| _version_ | 1818879903023497216 |
|---|---|
| author | David Kreuzer Michael Munz Stephan Schlüter |
| author_facet | David Kreuzer Michael Munz Stephan Schlüter |
| author_sort | David Kreuzer |
| collection | DOAJ |
| description | Local temperature forecasts for horizons up to 24 h are required in many applications. A common method to generate such forecasts is the Seasonal Autoregressive Integrated Moving Average (SARIMA) model or, much simpler, the naïve forecast. In this paper, we test whether deep neural networks are able to improve on the results from the above mentioned methods. In addition to univariate long short-term memory (LSTM) networks, we present an alternative method based on a 2D-convolutional LSTM (convLSTM) network. For benchmarking our approach we set up a case study using data from five different weather stations in Germany. The SARIMA model and the univariate LSTM network perform quite well in the first few hours, but are then outperformed by the multivariate LSTM network and our convolutional LSTM network for longer forecast horizons. Besides, both multivariate approaches show better performance when the temperature is changing in the course of the day. Overall, our presented approach based on a convolutional LSTM network performs best on all used test data sets. |
| first_indexed | 2024-12-19T14:37:28Z |
| format | Article |
| id | doaj.art-383be625c127413e8f71f162ae76816d |
| institution | Directory Open Access Journal |
| issn | 2666-8270 |
| language | English |
| last_indexed | 2024-12-19T14:37:28Z |
| publishDate | 2020-12-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Machine Learning with Applications |
| spelling | doaj.art-383be625c127413e8f71f162ae76816d2022-12-21T20:17:12ZengElsevierMachine Learning with Applications2666-82702020-12-012100007Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in GermanyDavid Kreuzer0Michael Munz1Stephan Schlüter2Ulm University of Applied Sciences, Albert-Einstein-Allee 55, 89081 Ulm, GermanyUlm University of Applied Sciences, Albert-Einstein-Allee 55, 89081 Ulm, GermanyUlm University of Applied Sciences, Prittwitzstraße 10, 89075 Ulm, Germany; Corresponding author.Local temperature forecasts for horizons up to 24 h are required in many applications. A common method to generate such forecasts is the Seasonal Autoregressive Integrated Moving Average (SARIMA) model or, much simpler, the naïve forecast. In this paper, we test whether deep neural networks are able to improve on the results from the above mentioned methods. In addition to univariate long short-term memory (LSTM) networks, we present an alternative method based on a 2D-convolutional LSTM (convLSTM) network. For benchmarking our approach we set up a case study using data from five different weather stations in Germany. The SARIMA model and the univariate LSTM network perform quite well in the first few hours, but are then outperformed by the multivariate LSTM network and our convolutional LSTM network for longer forecast horizons. Besides, both multivariate approaches show better performance when the temperature is changing in the course of the day. Overall, our presented approach based on a convolutional LSTM network performs best on all used test data sets.http://www.sciencedirect.com/science/article/pii/S2666827020300074Deep learningLSTMSARIMATemperature forecasts |
| spellingShingle | David Kreuzer Michael Munz Stephan Schlüter Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany Machine Learning with Applications Deep learning LSTM SARIMA Temperature forecasts |
| title | Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany |
| title_full | Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany |
| title_fullStr | Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany |
| title_full_unstemmed | Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany |
| title_short | Short-term temperature forecasts using a convolutional neural network — An application to different weather stations in Germany |
| title_sort | short term temperature forecasts using a convolutional neural network an application to different weather stations in germany |
| topic | Deep learning LSTM SARIMA Temperature forecasts |
| url | http://www.sciencedirect.com/science/article/pii/S2666827020300074 |
| work_keys_str_mv | AT davidkreuzer shorttermtemperatureforecastsusingaconvolutionalneuralnetworkanapplicationtodifferentweatherstationsingermany AT michaelmunz shorttermtemperatureforecastsusingaconvolutionalneuralnetworkanapplicationtodifferentweatherstationsingermany AT stephanschluter shorttermtemperatureforecastsusingaconvolutionalneuralnetworkanapplicationtodifferentweatherstationsingermany |