Reconstructing coupled time series in climate systems using three kinds of machine-learning methods

Huang, Yu; Yang, Lichao; Fu, Zuntao

doi:10.5194/esd-11-835-2020

Cited by 16 publications

(9 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Normally, the Pearson feature selection is used for linear relationships, but a weak linear correlation does not mean that there is no coupling relation between the variables (Huang et al, 2020) used similar approach for complex models.…”

Section: Data Descriptionmentioning

confidence: 99%

Outdoor Relative Humidity Prediction via Machine Learning Techniques

Zarinkamar¹,

Mayorga

2021

J ENVIRON INFORM LET

View full text Add to dashboard Cite

In an environmental control system, relative humidity (RH) is a very important factor because of its direct impact on humans or even animals and plants. However, there are few studies focused on prediction of humidity variables. The main objective of this paper is to show the capability of machine learning algorithms for RH prediction. In this study, a Long-Short Term Memory (LSTM) and four popular machine learning algorithms; namely, Multi-Layer Perceptron (MLP), Random Forest (RF), k-Nearest Neighbor (KNN) and Support Vector Machine for Regression (SVMR) are presented for a year period of time-series relative humidity to predict for a particular case in an Italian city. In order to have precise performance, data pre-processing is done before running the models. This thorough examination proves the positive effect of all Machine Learning-based algorithms in time-series relative humidity prediction based on predictive accuracy. Over the different metrics, LSTM indicates the best performance among all considered algorithms.

show abstract

Section: Data Descriptionmentioning

confidence: 99%

Outdoor Relative Humidity Prediction via Machine Learning Techniques

Zarinkamar¹,

Mayorga

2021

J ENVIRON INFORM LET

View full text Add to dashboard Cite

show abstract

“…However, links between climate fields can be nonlinear (Schneider et al, 2018;Dueben and Bauer, 2018;Huntingford et al, 2019;Nadiga, 2020). Nonlinear machine-learning-based CFR methods (for instance, artificial neural networks, ANN) could help capture underlying linear and nonlinear relationships between proxy records and the large-scale climate (Rasp and Lerch, 2018;Schneider et al, 2018;Rolnick et al, 2019;Huang et al, 2020;Nadiga, 2020;Chattopadhyay et al, 2020;Lindgren et al, 2021). Moreover, machine-learning methods do not necessarily rely on statistical methods to first obtain the principal spatial climate patterns, such as principal component analysis (PCA).…”

Section: Introductionmentioning

confidence: 99%

Evaluation of statistical climate reconstruction methods based on pseudoproxy experiments using linear and machine-learning methods

et al. 2022

View full text Add to dashboard Cite

Abstract. Three different climate field reconstruction (CFR) methods are employed to reconstruct spatially resolved North Atlantic–European (NAE) and Northern Hemisphere (NH) summer temperatures over the past millennium from proxy records. These are tested in the framework of pseudoproxy experiments derived from two climate simulations with comprehensive Earth system models. Two of these methods are traditional multivariate linear methods (principal component regression, PCR, and canonical correlation analysis, CCA), whereas the third method (bidirectional long short-term memory neural network, Bi-LSTM) belongs to the category of machine-learning methods. In contrast to PCR and CCA, Bi-LSTM does not need to assume a linear and temporally stable relationship between the underlying proxy network and the target climate field. In addition, Bi-LSTM naturally incorporates information about the serial correlation of the time series. Our working hypothesis is that the Bi-LSTM method will achieve a better reconstruction of the amplitude of past temperature variability. In all tests, the calibration period was set to the observational period, while the validation period was set to the pre-industrial centuries. All three methods tested herein achieve reasonable reconstruction performance on both spatial and temporal scales, with the exception of an overestimation of the interannual variance by PCR, which may be due to overfitting resulting from the rather short length of the calibration period and the large number of predictors. Generally, the reconstruction skill is higher in regions with denser proxy coverage, but it is also reasonably high in proxy-free areas due to climate teleconnections. All three CFR methodologies generally tend to more strongly underestimate the variability of spatially averaged temperature indices as more noise is introduced into the pseudoproxies. The Bi-LSTM method tested in our experiments using a limited calibration dataset shows relatively worse reconstruction skills compared to PCR and CCA, and therefore our working hypothesis that a more complex machine-learning method would provide better reconstructions for temperature fields was not confirmed. In this particular application with pseudoproxies, the implied link between proxies and climate fields is probably close to linear. However, a certain degree of reconstruction performance achieved by the nonlinear LSTM method shows that skill can be achieved even when using small samples with limited datasets, which indicates that Bi-LSTM can be a tool for exploring the suitability of nonlinear CFRs, especially in small data regimes.

show abstract

“…However, climate change is dynamic and chaotic, and many links between climate fields can be non-linear (Schneider et al, 2018;Dueben and Bauer, 2018;Huntingford et al, 2019;Nadiga, 2020). Nonlinear machine leaning-based CFR methods (for instance, Artificial Neural Networks-ANN) could help capture capture underlying linear and nonlinear relationships between proxy records and the large-scale climate as realistically as possible (Rasp and Lerch, 2018;Schneider et al, 2018;Rolnick et al, 2019;Huang et al, 2020;Nadiga, 2020;Chattopadhyay et al, 2020;Lindgren et al, 2021). Moreover, machine-learning methods do not necessarily rely on statistical methods to first obtain the principal spatial climate patterns, such as Principal Component Analysis-PCA.…”

mentioning

confidence: 99%

“…The state of the output gate ot is calculated from the previous hidden state and the current input variables (equation 11). This output is used to compute the updated hidden state ht using the state of the cell output Ct (equation 12)(Huang et al, 2020;Chattopadhyay et al, 2020).In the present application to climate reconstructions, we have a set of input pseudoproxy data𝑿 𝑡 𝑛 = [xt-i,…, xt-1]and an output target temperature time series 𝒀 𝑡 𝑚 = [yt-i,…, yt-1]. The forward LSTM hidden state sequence 𝒉 𝑡 ⃑⃑⃑⃑ (note the arrow direction) is calculated employing inputs information in a positive direction from time t-1 to time t-n iteratively, and for backward LSTM…”

mentioning

confidence: 99%

Evaluation of statistical climate reconstruction methods based on pseudoproxy experiments using linear and machine learning methods

Zhang

Wagner

Klockmann

et al. 2022

Preprint

View full text Add to dashboard Cite

Abstract. Three different climate field reconstruction (CFR) methods employed to reconstruct North Atlantic-European (NAE) and Northern Hemisphere (NH) summer season temperature over the past millennium from proxy records are tested in the framework of pseudoproxy experiments derived from three climate simulations with Earth System Models. Two of these methods are traditional multivariate linear methods (Principal Components Regression, PCR and Canonical Correlation Analysis, CCA), whereas the third method (Bidirectional Long-Short-Term Memory Neural Network, Bi-LSTM) belongs to the category of machine learning methods. The Bi-LSTM method does not need to assume linear and temporally stable relationships between the underlying proxy network and the targeted climate field, in contrast to PCR and CCA. In addition, Bi-LSTM incorporates information on the serial correlation of the time series. All three methods tested herein achieve reasonable reconstruction performance in both spatial and temporal scale. Generally, the reconstruction skill is higher in regions with denser proxy coverage, but reconstruction skill is also achieved in proxy-free areas due to climate teleconnections. All three CFR methodologies generally tend to more strongly underestimate the target temperature variations as more noise is introduced into the pseudoproxies. The Bi-LSTM method tested in our experiments shows relatively worse reconstruction skills compared to PCR and CCA, yet it brings some encouraging results on capturing extreme cooling climate signals. This indicates that this nonlinear CFR method could be a potential methodology for past climate extremes analysis.

show abstract

Reconstructing coupled time series in climate systems using three kinds of machine-learning methods

Cited by 16 publications

References 61 publications

Outdoor Relative Humidity Prediction via Machine Learning Techniques

Outdoor Relative Humidity Prediction via Machine Learning Techniques

Evaluation of statistical climate reconstruction methods based on pseudoproxy experiments using linear and machine-learning methods

Evaluation of statistical climate reconstruction methods based on pseudoproxy experiments using linear and machine learning methods

Contact Info

Product

Resources

About