DLENSO: A Deep Learning ENSO Forecasting Model

He, Dandan; Lin, Pengfei; Liu, Hailong; Ding, Lei; Jiang, Junguang

doi:10.1007/978-3-030-29911-8_2

Cited by 22 publications

(18 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(Though the output of their model is still the Niño 3.4 index, they construct the model and make forecasts by absorbing the historical spatial-temporal features from variable patterns instead of previous index records, so we mark this study as SST pattern forecasts in this paper.) Mu et al (2019) and He et al (2019) built a ConvLSTM (Shi et al, 2015) model to capture the spatial-temporal dependencies of ENSO SST patterns over multiple time horizons and obtained better predictions. Zheng et al (2020) constructed a purely satellite-data-driven deep learning model to forecast the evolutions of tropical instability wave, which is closely related to ENSO phenomena, and obtained accurate and efficient forecasts.…”

Section: Introductionmentioning

confidence: 99%

ENSO-ASC 1.0.0: ENSO deep learning forecast model with a multivariate air–sea coupler

Qin

Yuan

2021

Geosci. Model Dev.

View full text Add to dashboard Cite

Abstract. The El Niño–Southern Oscillation (ENSO) is an extremely complicated ocean–atmosphere coupling event, the development and decay of which are usually modulated by the energy interactions between multiple physical variables. In this paper, we design a multivariate air–sea coupler (ASC) based on the graph using features of multiple physical variables. On the basis of this coupler, an ENSO deep learning forecast model (named ENSO-ASC) is proposed, whose structure is adapted to the characteristics of the ENSO dynamics, including the encoder and decoder for capturing and restoring the multi-scale spatial–temporal correlations, and two attention weights for grasping the different air–sea coupling strengths on different start calendar months and varied effects of physical variables in ENSO amplitudes. In addition, two datasets modulated to the same resolutions are used to train the model. We firstly tune the model performance to optimal and compare it with the other state-of-the-art ENSO deep learning forecast models. Then, we evaluate the ENSO forecast skill from the contributions of different predictors, the effective lead time with different start calendar months, and the forecast spatial uncertainties, to further analyze the underlying ENSO mechanisms. Finally, we make ENSO predictions over the validation period from 2014 to 2020. Experiment results demonstrate that ENSO-ASC outperforms the other models. Sea surface temperature (SST) and zonal wind are two crucial predictors. The correlation skill of the Niño 3.4 index is over 0.78, 0.65, and 0.5 within the lead time of 6, 12, and 18 months respectively. From two heat map analyses, we also discover the common challenges in ENSO predictability, such as the forecasting skills declining faster when making forecasts through June–July–August and the forecast errors being more likely to show up in the western and central tropical Pacific Ocean in longer-term forecasts. ENSO-ASC can simulate ENSO with different strengths, and the forecasted SST and wind patterns reflect an obvious Bjerknes positive feedback mechanism. These results indicate the effectiveness and superiority of our model with the multivariate air–sea coupler in predicting ENSO and analyzing the underlying dynamic mechanisms in a sophisticated way.

show abstract

Section: Introductionmentioning

confidence: 99%

ENSO-ASC 1.0.0: ENSO deep learning forecast model with a multivariate air–sea coupler

Qin

Yuan

2021

Geosci. Model Dev.

View full text Add to dashboard Cite

show abstract

“…To mitigate the drawbacks of the LSTM model, Xingjian et al (2015) proposed a convolution LSTM (ConvLSTM) architecture to implement the precipitation prediction, where convolution layers are added based on LSTM to capture spatial features. He et al (2019) proposed a DLENSO model based on ConvLSTM to forecast ENSO events, and the simulation results indicate that it outperforms the conventional LSTM model. Gupta et al (2020) proved that using a ConvLSTM network to predict the Niño3.4 index overcomes the spring predictability barrier.…”

Section: Introductionmentioning

confidence: 99%

Multi-Year ENSO Forecasts Using Parallel Convolutional Neural Networks With Heterogeneous Architecture

Nie

Liu

et al. 2021

Front. Mar. Sci.

View full text Add to dashboard Cite

The El Niño-Southern Oscillation (ENSO) is one of the main drivers of the interannual climate variability of Earth and can cause a wide range of climate anomalies, so multi year ENSO forecasts are a paramount scientific issue. However, most existing works rely on the conventional iterative mechanism and, thus, fail to provide reliable long-term predictions due to error accumulation. Although methods based on deep learning (DL) apply the parallel modeling scheme for different lead times instead of a single iteration model, they leverage the same DL model for prediction, which can not fully mine the variability of different lead times, resulting in a decrease of prediction accuracy. To solve this problem, we propose a novel parallel deep convolutional neural network (CNN) with a heterogeneous architecture. In this study, by adaptively selecting network architectures for different lead times, we realize variability modeling of different tasks (lead times) and thereby improve the reliability of long-term predictions. Furthermore, we propose a relationship between different prediction lead times and neural network architecture from a unique perspective, namely, the receptive field originally proposed in computer vision. According to the spatio-temporal correlated area and sampling scale of lead times, the size of the convolution kernel and the mesh size of sampling are adjusted as the lead time increases. The Coupled Model Intercomparison Project phase 5 (CMIP5) from 1861 to 2004 and the Simple Ocean Data Assimilation (SODA) from 1871 to 1973 were used for model training, and the GODAS from 1982 to 2017 were used for testing the forecast skill of the model. Experimental results demonstrate that the proposed method outperforms the other well-known methods, especially for long-term predictions.

show abstract

“…Ham et al (2019) applies transfer learning (Yosinski et al, 2014) on historical simulations from CMIP5 (Coupled Model Intercomparison Project phase5, Bellenger et al, 2014) and reanalysis data with a CNN model to predict ENSO events, resulting in a robust and long-term forecast for up to 1.5 years, which outperforms the current numerical predictions. Mu et al (2019) and He et al (2019) both build a ConvLSTM (Shi et al, 2015) model to capture the spatial-temporal dependencies of ENSO SST patterns over multiple time horizons and obtain better predictions. Zheng et al (2020) constructs a purely satellite data-driven deep learning to forecast the evolutions of tropical instability wave, which is closely related to ENSO phenomena, and obtain accurate and efficient forecasts.…”

Section: Introductionmentioning

confidence: 99%

ENSO-ASC 1.0.0: ENSO Deep Learning Forecast Model with a Multivariate Air–Sea Coupler

Mu¹,

Qin²,

Yuan³

2021

Preprint

View full text Add to dashboard Cite

Abstract. ENSO is an extremely sophisticated air-sea coupling phenomenon, the development and decay of which are usually modulated by the energy interactions between multiple physical variables. In this paper, we design a multivariate air-sea coupler (ASC) based on graph using features of multiple physical variables. On the basis of the coupler, an ENSO deep learning forecast model (named ENSO-ASC) is proposed, whose structure is adapted to the characteristics of the ENSO dynamics, including the encoder/decoder for capturing/restoring the multi-scale spatial-temporal correlations, and two attention components for grasping the different air-sea coupling strength on different start calendar month and varied contributions of physical variables in ENSO amplitudes. In addition, two datasets at different resolutions are used to train the model. We firstly tune the model performance to optimal and compare it with the other state-of-the-art ENSO deep learning forecast models. Then, we evaluate the ENSO forecast skill from the contributions of different predictors, the effective lead time with the different start calendar months, and the forecast spatial uncertainties, further analyze the underlying ENSO mechanisms. Finally, we make ENSO predictions over the validation period from 2014 to 2020. Experiment results demonstrate that ENSO-ASC outperforms the other models. Sea surface temperature (SST) and zonal wind are two crucial predictors. The correlation skill of Niño3.4 index is over 0.78/0.65/0.5 within the lead time of 6/12/18 months. From two heat map analyses, we also discover the common challenges in ENSO predictability, such as the forecasting skills declining faster when making forecasts through June-July-August and the forecast errors more likely showing up in the western-central equatorial Pacific with a longer lead time. ENSO-ASC can simulate El Niño and La Niña events with different strengths. The forecasted SST and wind patterns reflect obvious Bjerknes positive feedback mechanism. These results indicate the eﬀectiveness and superiority of our model with the multivariate air-sea coupler in predicting sophisticated ENSO and analyzing the underlying dynamic mechanisms.

show abstract

DLENSO: A Deep Learning ENSO Forecasting Model

Cited by 22 publications

References 15 publications

ENSO-ASC 1.0.0: ENSO deep learning forecast model with a multivariate air–sea coupler

ENSO-ASC 1.0.0: ENSO deep learning forecast model with a multivariate air–sea coupler

Multi-Year ENSO Forecasts Using Parallel Convolutional Neural Networks With Heterogeneous Architecture

ENSO-ASC 1.0.0: ENSO Deep Learning Forecast Model with a Multivariate Air–Sea Coupler

Contact Info

Product

Resources

About