Performance of statistical and machine learning ensembles for daily temperature downscaling

Temperature changes have widespread impacts on the environment, economy, and municipal planning. Generating accurate climate prediction at finer spatial resolution through downscaling could help better assess the future effects of climate change on a local scale. Ensembles of multiple climate models have been proven to improve the accuracy of temperature prediction. Meanwhile, machine learning techniques have shown high performance in solving various predictive modeling problems, which make them a promising tool for temperature downscaling. This study investigated the performance of machine learning (long short-term memory (LSTM) networks and support vector machine (SVM)) and statistical (arithmetic ensemble mean (EM) and multiple linear regression (MLR)) methods in developing multi-model ensembles for downscaling long-term daily temperature. A case study of twelve meteorological stations across Ontario, Canada, was conducted to evaluate the performance of the proposed ensembles. The results showed that both machine learning and statistical techniques performed well at downscaling daily temperature with multi-model ensembles and had similar performance with relatively high accuracy. The R 2 of 12 stations ranged between 0.756 and 0.820 and RMSE ranged between 4.318 and 7.063 °C. Both machine learning and statistical ensembles for downscaling had difficulty in predicting extreme values for temperature below − 10 °C and above 20 °C. The results provided technical support for using statistical and machine learning methods to generate high-resolution daily temperature prediction.

[1]  Ravi S. Nanjundiah,et al.  Role of predictors in downscaling surface temperature to river basin in India for IPCC SRES scenarios using support vector machine , 2009 .

[2]  Zhong Li,et al.  A stepwise cluster analysis approach for downscaled climate projection - A Canadian case study , 2013, Environ. Model. Softw..

[3]  Ajith H. Perera,et al.  Ecology of a managed terrestrial landscape : patterns and processes of forest landscapes in Ontario , 2000 .

[4]  Yinhai Wang,et al.  Short-Term Speed Prediction Using Remote Microwave Sensor Data: Machine Learning versus Statistical Model , 2016 .

[5]  Alex J. Cannon,et al.  Evaluation of Linear and Non-Linear Downscaling Methods in Terms of Daily Variability and Climate Indices: Surface Temperature in Southern Ontario and Quebec, Canada , 2014 .

[6]  J. Prueger,et al.  Temperature extremes: Effect on plant growth and development , 2015 .

[7]  Shamsuddin Shahid,et al.  Model output statistics downscaling using support vector machine for the projection of spatial and temporal changes in rainfall of Bangladesh , 2018, Atmospheric Research.

[8]  Xue-Jie Gao,et al.  Building Asian climate change scenario by multi‐regional climate models ensemble. Part I: surface air temperature , 2016 .

[9]  Richard Betts,et al.  Implications of climate change for agricultural productivity in the early twenty-first century , 2010, Philosophical Transactions of the Royal Society B: Biological Sciences.

[10]  Paul Chinowsky,et al.  Climate change risks to US infrastructure: impacts on roads, bridges, coastal development, and urban drainage , 2015, Climatic Change.

[11]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[12]  Bin Wang,et al.  Using multi‐model ensembles of CMIP5 global climate models to reproduce observed monthly rainfall and temperature with machine learning methods in Australia , 2018, International Journal of Climatology.

[13]  O. Kisi,et al.  Modelling long‐term monthly temperatures by several data‐driven methods using geographical inputs , 2015 .

[14]  Kang-Kun Lee,et al.  A method to improve the stability and accuracy of ANN- and SVM-based time series models for long-term groundwater level predictions , 2016, Computational Geosciences.

[15]  Afan Galih Salman,et al.  Single Layer & Multi-layer Long Short-Term Memory (LSTM) Model with Intermediate Variables for Weather Forecasting , 2018 .

[16]  F. Giorgi,et al.  Addressing climate information needs at the regional level: the CORDEX framework , 2009 .

[17]  Kamal Ahmed,et al.  Statistical downscaling of precipitation using machine learning techniques , 2018, Atmospheric Research.

[18]  Quanqin Shao,et al.  Using multi-model ensembles to improve the simulated effects of land use/cover change on temperature: a case study over northeast China , 2016, Climate Dynamics.

[19]  Sa-Kwang Song,et al.  DeepRain: ConvLSTM Network for Precipitation Prediction using Multichannel Radar Data , 2017, ArXiv.

[20]  Hisashi Kato,et al.  Comparison of four ensemble methods combining regional climate simulations over Asia , 2011 .

[21]  B. Baetz,et al.  Performance of multi-model ensembles for the simulation of temperature variability over Ontario, Canada , 2018, Environmental Earth Sciences.

[22]  Alexander J. Smola,et al.  Support Vector Regression Machines , 1996, NIPS.

[23]  Erik Kjellström,et al.  Using and Designing GCM–RCM Ensemble Regional Climate Projections , 2010 .

[24]  Hyun-Han Kwon,et al.  A modified support vector machine based prediction model on streamflow at the Shihmen Reservoir, Taiwan , 2010 .

[25]  Frédéric Magoulès,et al.  A review on the prediction of building energy consumption , 2012 .

[26]  H. Cai,et al.  Evaluation of SVM, ELM and four tree-based ensemble models for predicting daily reference evapotranspiration using limited meteorological data in different climates of China , 2018, Agricultural and Forest Meteorology.

[27]  Thanh Duc Dang,et al.  Downscaling rainfall using deep learning long short‐term memory and feedforward neural network , 2019, International Journal of Climatology.

[28]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[29]  Alex Graves,et al.  Generating Sequences With Recurrent Neural Networks , 2013, ArXiv.

[30]  T. N. Krishnamurti,et al.  Experimental real-time multi-model ensemble (MME) prediction of rainfall during monsoon 2008: Large-scale medium-range aspects , 2011 .

[31]  T. N. Krishnamurti,et al.  Improving Global Model Precipitation Forecasts over India Using Downscaling and the FSU Superensemble. Part I: 1–5-Day Forecasts , 2009 .

[32]  William W. Hsieh,et al.  Forecasting regional sea surface temperatures in the tropical Pacific by neural network models, with wind stress and sea level pressure as predictors , 1998 .

[33]  S. Chalov,et al.  Surface water pathways and fluxes of metals under changing environmental conditions and human interventions in the Selenga River system , 2016, Environmental Earth Sciences.

[34]  V. V. Srinivas,et al.  Downscaling of precipitation for climate change scenarios: A support vector machine approach , 2006 .

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

[36]  Y. Radhika,et al.  Atmospheric Temperature Prediction using Support Vector Machines , 2009 .

[37]  Karsten Schulz,et al.  Rainfall–runoff modelling using Long Short-Term Memory (LSTM) networks , 2018, Hydrology and Earth System Sciences.

[38]  Evangelos Spiliotis,et al.  Statistical and Machine Learning forecasting methods: Concerns and ways forward , 2018, PloS one.

[39]  S. Somot,et al.  Evaluation of the regional climate model ALADIN to simulate the climate over North America in the CORDEX framework , 2013, Climate Dynamics.

[40]  Chengyi Wang,et al.  Long Short-Term Memory Neural Networks for Online Disturbance Detection in Satellite Image Time Series , 2018, Remote. Sens..

[41]  F. Giorgi,et al.  Weight assignment in regional climate models , 2010 .

[42]  Kevin M. Smith,et al.  Forecasting energy consumption of multi-family residential buildings using support vector regression: Investigating the impact of temporal and spatial monitoring granularity on performance accuracy , 2014 .

[43]  M. Ye,et al.  Developing a Long Short-Term Memory (LSTM) based model for predicting water table depth in agricultural areas , 2018, Journal of Hydrology.

[44]  Yuqing Wang,et al.  Building Asian climate change scenario by multi‐regional climate models ensemble. Part II: mean precipitation , 2016 .

[45]  Vladimir M. Krasnopolsky,et al.  A Neural Network Nonlinear Multimodel Ensemble to Improve Precipitation Forecasts over Continental US , 2012 .

[46]  I. D. Gates,et al.  Support vector regression for porosity prediction in a heterogeneous reservoir: A comparative study , 2010, Comput. Geosci..

[47]  J. Jarsjö,et al.  Climate-driven change of nitrogen retention–attenuation near irrigated fields: multi-model projections for Central Asia , 2017, Environmental Earth Sciences.

[48]  Guohe Huang,et al.  Ensemble Projections of Regional Climatic Changes over Ontario, Canada , 2015 .

[49]  A. K. Gosain,et al.  Dynamic coupling of support vector machine and K-nearest neighbour for downscaling daily rainfall , 2015 .

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

[51]  Xiao Liu,et al.  Intelligent initial map scale generation based on rough-set rules , 2019, Arabian Journal of Geosciences.

[52]  Steve M. Easterbrook,et al.  A Machine Learning Approach to Non-uniform Spatial Downscaling of Climate Variables , 2017, 2017 IEEE International Conference on Data Mining Workshops (ICDMW).

[53]  T. B. M. J. Ouarda,et al.  Comparison of transfer functions in statistical downscaling models for daily temperature and precipitation over Canada , 2012, Stochastic Environmental Research and Risk Assessment.

[54]  Renate Hagedorn,et al.  The rationale behind the success of multi-model ensembles in seasonal forecasting — I. Basic concept , 2005 .

[55]  Paulin Coulibaly,et al.  Temporal neural networks for downscaling climate variability and extremes , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[56]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[57]  A. Massi Pavan,et al.  Least squares support vector machine for short-term prediction of meteorological time series , 2012, Theoretical and Applied Climatology.

[58]  Alex de A. Fernandes,et al.  Multi-model ensemble: technique and validation , 2014 .

[59]  Ozgur Kisi,et al.  Prediction of long‐term monthly precipitation using several soft computing methods without climatic data , 2015 .

[60]  T. N. Krishnamurti,et al.  Improving Global Model Precipitation Forecasts over India Using Downscaling and the FSU Superensemble. Part II: Seasonal Climate , 2009 .

[61]  A. Bohra,et al.  Multi‐model ensemble (MME) prediction of rainfall using neural networks during monsoon season in India , 2012 .

[62]  P. Samui.  Slope stability analysis: a support vector machine approach , 2008 .

[63]  Trevor Darrell,et al.  Long-term recurrent convolutional networks for visual recognition and description , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[64]  Song Li,et al.  Weather Forecasting Using Artificial Neural Network , 2019, ICAIS.

[65]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[66]  Quoc V. Le,et al.  Sequence to Sequence Learning with Neural Networks , 2014, NIPS.

[67]  Mikhail Chester,et al.  Impacts of climate change on electric power supply in the Western United States , 2015 .

[68]  T. N. Krishnamurti,et al.  Improved Weather and Seasonal Climate Forecasts from Multimodel Superensemble. , 1999, Science.

[69]  Ramesh Sharda,et al.  Connectionist approach to time series prediction: an empirical test , 1992, J. Intell. Manuf..

[70]  X. Y. Li,et al.  Prediction of Long-Term Near-Surface Temperature Based on NA-CORDEX Output , 2019, Journal of Environmental Informatics Letters.

[71]  C. M. Kishtawal,et al.  Multimodel Ensemble Forecasts for Weather and Seasonal Climate , 2000 .

[72]  Yingchun Ren,et al.  Sparsity Preserving Discriminant Projections with Applications to Face Recognition , 2016 .

[73]  M. A. Rafe Biswas,et al.  Regression analysis for prediction of residential energy consumption , 2015 .

[74]  P. Coulibaly,et al.  Downscaling Precipitation and Temperature with Temporal Neural Networks , 2005 .

[75]  Reimund P. Rötter,et al.  Lessons from climate modeling on the design and use of ensembles for crop modeling , 2016, Climatic Change.

[76]  Jianhua Xu,et al.  Modeling the multiple time scale response of hydrological drought to climate change in the data-scarce inland river basin of Northwest China , 2019, Arabian Journal of Geosciences.

[77]  Ning Xu,et al.  Forecasting district-scale energy dynamics through integrating building network and long short-term memory learning algorithm , 2019, Applied Energy.

[78]  M. Kadioǧlu,et al.  Multischeme ensemble forecasting of surface temperature using neural network over Turkey , 2012, Theoretical and Applied Climatology.

[79]  G. Huang,et al.  High-Resolution Probabilistic Projections of Temperature Changes over Ontario, Canada , 2014 .