An Ensemble 3D Convolutional Neural Network for Spatiotemporal Soil Temperature Forecasting

Soil temperature (ST) plays an important role in agriculture and other fields, and has a close relationship with plant growth and development. Therefore, accurate ST prediction methods are widely needed. Deep learning (DL) models have been widely applied for ST prediction. However, the traditional DL models may fail to capture the spatiotemporal relationship due to its complex dependency under different related hydrologic variables. Hence, the DL models with Ensemble Empirical Mode Decomposition (EEMD) are proposed in this study. The proposed models can capture more complex spatiotemporal relationship after decomposing the ST into different intrinsic mode functions. Therefore, the performance of models is further improved. The results show that the performance of DL models with EEMD are better than that of corresponding DL models without EEMD. Moreover, EEMD-Conv3d has the best performance among all the experimental models. It has the highest R2 ranging from 0.9826 to 0.9893, the lowest RMSE ranging from 1.3096 to 1.6497 and the lowest MAE ranging from 0.9656 to 1.2056 in predicting ST at the lead time from one to five days. In addition, the lines between predicted ST and observed ST are closer to the ideal line (y = x) than other DL models. The results show that our EEMD-Conv3D can better capture spatiotemporal correlation and is an applicable method for predicting spatiotemporal ST.

[1]  Lorenzo Torresani,et al.  Learning Spatiotemporal Features with 3D Convolutional Networks , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[2]  Clive W. J. Granger,et al.  Non-linear time series modeling , 2001 .

[3]  Q. Tan,et al.  An adaptive middle and long-term runoff forecast model using EEMD-ANN hybrid approach , 2018, Journal of Hydrology.

[4]  Gozde Unal,et al.  A RNN based time series approach for forecasting turkish electricity load , 2018 .

[5]  Pradeep Hewage,et al.  Long-Short Term Memory for an Effective Short-Term Weather Forecasting Model Using Surface Weather Data , 2019, AIAI.

[6]  G. Mihalakakou,et al.  On estimating soil surface temperature profiles , 2002 .

[7]  P. Willems,et al.  Short‐term forecasting of soil temperature using artificial neural network , 2015 .

[8]  Ehsan Mohammadi,et al.  Modeling daily soil temperature over diverse climate conditions in Iran—a comparison of multiple linear regression and support vector regression techniques , 2019, Theoretical and Applied Climatology.

[9]  Wojciech Zaremba,et al.  Recurrent Neural Network Regularization , 2014, ArXiv.

[10]  M. Bilgili.  Prediction of soil temperature using regression and artificial neural network models , 2010 .

[11]  Ce Liu,et al.  Deep Convolutional Neural Network for Image Deconvolution , 2014, NIPS.

[12]  Jia-Qi Zhu,et al.  Improved EEMD-based crude oil price forecasting using LSTM networks , 2019, Physica A: Statistical Mechanics and its Applications.

[13]  Dit-Yan Yeung,et al.  Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting , 2015, NIPS.

[14]  Keryn I. Paul,et al.  Soil temperature under forests: a simple model for predicting soil temperature under a range of forest types , 2004 .

[15]  Pengjian Shang,et al.  Multidimensional k-nearest neighbor model based on EEMD for financial time series forecasting , 2017 .

[16]  Uri Weiser,et al.  Spatial Correlation and Value Prediction in Convolutional Neural Networks , 2018, IEEE Computer Architecture Letters.

[17]  Soukayna Mouatadid,et al.  WeatherBench: A Benchmark Data Set for Data‐Driven Weather Forecasting , 2020, Journal of Advances in Modeling Earth Systems.

[18]  L. Liang,et al.  A simple framework to estimate distributed soil temperature from discrete air temperature measurements in data‐scarce regions , 2014 .

[19]  Ming Yang,et al.  3D Convolutional Neural Networks for Human Action Recognition , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[20]  Chao-Ming Huang,et al.  Analysis of an adaptive time-series autoregressive moving-average (ARMA) model for short-term load forecasting , 1995 .

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

[22]  Qiuwen Zhang,et al.  A Novel Hybrid Data-Driven Model for Daily Land Surface Temperature Forecasting Using Long Short-Term Memory Neural Network Based on Ensemble Empirical Mode Decomposition , 2018, International journal of environmental research and public health.

[23]  Ah Chung Tsoi,et al.  Face recognition: a convolutional neural-network approach , 1997, IEEE Trans. Neural Networks.

[24]  Rattan Lal,et al.  Diurnal soil temperature fluctuations for different erosion classes of an oxisol at Mlingano, Tanzania , 1998 .

[25]  Georg Dorffner,et al.  Neural Networks for Time Series Processing , 1996 .

[26]  A. Moosavi,et al.  Time series modelling of increased soil temperature anomalies during long period , 2015 .

[27]  Norden E. Huang,et al.  Ensemble Empirical Mode Decomposition: a Noise-Assisted Data Analysis Method , 2009, Adv. Data Sci. Adapt. Anal..

[28]  Kwok-wing Chau,et al.  Improving Forecasting Accuracy of Annual Runoff Time Series Using ARIMA Based on EEMD Decomposition , 2015, Water Resources Management.

[29]  R. E. Abdel-Aal,et al.  Hourly temperature forecasting using abductive networks , 2004, Eng. Appl. Artif. Intell..

[30]  Ozgur Kisi,et al.  Soil temperature modeling at different depths using neuro-fuzzy, neural network, and genetic programming techniques , 2017, Theoretical and Applied Climatology.

[31]  Demetris F. Lekkas,et al.  APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR FLOOD FORECASTING , 2004 .

[32]  Donald F. Specht,et al.  A general regression neural network , 1991, IEEE Trans. Neural Networks.

[33]  A. Harvey,et al.  Forecasting Economic Time Series With Structural and Box-Jenkins Models: A Case Study , 1983 .

[34]  X. Chen,et al.  Soil Respiration in Different Agricultural and Natural Ecosystems in an Arid Region , 2012, PloS one.

[35]  Yu Zhang,et al.  Soil temperature in Canada during the twentieth century: Complex responses to atmospheric climate change , 2005 .

[36]  N. Huang,et al.  The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis , 1998, Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.

[37]  Phil Blunsom,et al.  A Convolutional Neural Network for Modelling Sentences , 2014, ACL.

[38]  N. Huang,et al.  A new view of nonlinear water waves: the Hilbert spectrum , 1999 .

[39]  Brownmang onwuka,et al.  Effects of soil temperature on some soil properties and plant growth , 2018 .

[40]  B. Saavedra-Moreno,et al.  Monthly prediction of air temperature in Australia and New Zealand with machine learning algorithms , 2016, Theoretical and Applied Climatology.

[41]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.