Oceanic Harbingers of Pacific Decadal Oscillation Predictability in CESM2 Detected by Neural Networks

Gordon, Emily M; Barnes, Elizabeth A.; Hurrell, James W.

doi:10.1029/2021gl095392

Cited by 31 publications

(43 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Future work could explore the predictability of slowdowns using ANNs with other climate predictors, such as considering a TOA energy imbalance approach (Hedemann et al, 2017), or taking into account longer duration events. It may also be valuable to combine maps of OHC at different lead times, which was recently demonstrated by Gordon et al (2021) for predicting transitions in the phase of the PDO. Lastly, we train our ANN on a large ensemble from only one climate model.…”

Section: Discussionmentioning

confidence: 99%

“…To attempt to understand the ANN's decision‐making process, we use a method of XAI called layer‐wise relevance propagation (LRP; Bach et al., 2015; Montavon et al., 2017, 2018). The utility of LRP has been demonstrated in a wide range of weather and climate applications (e.g., Barnes, Toms, et al., 2020; Davenport & Diffenbaugh, 2021; Gordon et al., 2021; Labe & Barnes, 2021; Sonnewald & Lguensat, 2021), and an overview for the geosciences can be found in Toms et al. (2020).…”

Section: Methodsmentioning

confidence: 99%

“…Deep learning methods, such as neural networks, have the ability to extract and leverage nonlinear patterns across data‐intensive spatial fields, which make them promising tools for revealing new insights and sources of predictability in climate science (Barnes, Mayer, et al., 2020; Irrgang et al., 2021; Reichstein et al., 2019; Sonnewald et al., 2021). Recent work has demonstrated the utility for neural networks in identifying climate modes, teleconnections, and forecasts of opportunity for a wide variety of timescales (e.g., Gibson et al., 2021; Gordon et al., 2021; Ham et al., 2019; J. Liu et al., 2021; Mayer & Barnes, 2021; Nadiga, 2021; Tang & Duan, 2021; Toms et al., 2021; Wu & Hsieh, 2004). Further, a growing number of explainable artificial intelligence (XAI) methods have been adapted for applications in weather and climate science (McGovern et al., 2019; Toms et al., 2020), which can retrospectively trace the decisions of neural networks and assist scientists in comparing the attribution of input features to known physical mechanisms in the Earth system.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Predicting Slowdowns in Decadal Climate Warming Trends With Explainable Neural Networks

Labe

Barnes

2022

Geophysical Research Letters

Self Cite

View full text Add to dashboard Cite

The global mean surface temperature (GMST) record exhibits both interannual to multidecadal variability and a long‐term warming trend due to external climate forcing. To explore the predictability of temporary slowdowns in decadal warming, we apply an artificial neural network (ANN) to climate model data from the Community Earth System Model Version 2 Large Ensemble. Here, an ANN is tasked with whether or not there will be a slowdown in the rate of the GMST trend by using maps of ocean heat content (OHC) at the onset. Through a machine learning explainability method, we find the ANN is learning off‐equatorial patterns of anomalous OHC that resemble transitions in the phase of the Interdecadal Pacific Oscillation in order to make slowdown predictions. Finally, we test our ANN on observed historical data, which further reveals how explainable neural networks are useful tools for understanding decadal variability in both climate models and observations.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting Slowdowns in Decadal Climate Warming Trends With Explainable Neural Networks

Labe

Barnes

2022

Geophysical Research Letters

Self Cite

View full text Add to dashboard Cite

show abstract

“…This leaves exploration of the CAN's utility in specific scientific contexts to future research. For example, previous work using neural networks to identify climate forecasts of opportunity on subseasonal-to-decadal timescales could be extended by taking an abstention approach (e.g., Barnes, Mayer, et al, 2020;Gordon et al, 2021;Mayer & Barnes, 2021;Tseng et al, 2020).…”

Section: Discussionmentioning

confidence: 99%

Controlled Abstention Neural Networks for Identifying Skillful Predictions for Regression Problems

Barnes

2021

J Adv Model Earth Syst

Self Cite

View full text Add to dashboard Cite

The earth system is exceedingly complex and often chaotic in nature, making prediction incredibly challenging: we cannot expect to make perfect predictions all of the time. Instead, we look for specific states of the system that lead to more predictable behavior than others, often termed “forecasts of opportunity.” When these opportunities are not present, scientists need prediction systems that are capable of saying “I don't know.” We introduce a novel loss function, termed “abstention loss,” that allows neural networks to identify forecasts of opportunity for regression problems. The abstention loss works by incorporating uncertainty in the network's prediction to identify the more confident samples and abstain (say “I don't know”) on the less confident samples. The abstention loss is designed to determine the optimal abstention fraction, or abstain on a user‐defined fraction using a standard adaptive controller. Unlike many methods for attaching uncertainty to neural network predictions post‐training, the abstention loss is applied during training to preferentially learn from the more confident samples. The abstention loss is built upon nonlinear heteroscedastic regression, a standard computer science method. While nonlinear heteroscedastic regression is a simple yet powerful tool for incorporating uncertainty in regression problems, we demonstrate that the abstention loss outperforms it for the synthetic climate use cases explored here. The implementation of the proposed abstention loss is straightforward in most network architectures designed for regression, as it only requires modification of the output layer and loss function.

show abstract

“…The advent of big data has led to the successful application of deep learning methods to many predictive tasks. For example, artificial neural network (ANN) technology has been used to predict PDO persistence and transitions (Gordon et al., 2021), red tides (Qin et al., 2017) and air quality factors such as PM 2.5 (Kow et al., 2020). Convolutional neural network (CNN) technology, which possesses a powerful extraction ability for spatial features, has been used to model large‐scale sea surface temperature (SST) data and for forecasting the El Niño‐Southern Oscillation (ENSO) (Ham et al., 2019) and tropical instability waves (Zheng et al., 2020).…”

Section: Introductionmentioning

confidence: 99%

Deep Learning for Multi‐Timescales Pacific Decadal Oscillation Forecasting

Qin¹,

Du²,

Hu³

et al. 2022

Geophysical Research Letters

View full text Add to dashboard Cite

Variations in the Pacific decadal oscillation (PDO) can influence marine ecosystems and regional climate phenomena. Accurate and long‐term forecasts of the PDO are therefore crucial for marine governance. This paper presents a novel seasonal gated recurrent unit (SGRU) model, based on deep learning, for forecasting the PDO at multiple time scales. The model first decomposes the complex and nonlinear PDO index time series into three separate components, each retaining a distinct pattern of PDO. Next, a three‐pathway GRU model is constructed to model and forecast the PDO index. A test applying the SGRU model to the period from 1979 to 2020 demonstrates its superiority over eight state‐of‐the‐art models in PDO forecasting. Additionally, the SGRU model can flexibly produce high‐performance forecasts at multiple time scales. In view of that physical and dynamical models rely on clear evolutionary mechanisms, the SGRU model overcomes the complexities of these models.

show abstract

Oceanic Harbingers of Pacific Decadal Oscillation Predictability in CESM2 Detected by Neural Networks

Cited by 31 publications

References 45 publications

Predicting Slowdowns in Decadal Climate Warming Trends With Explainable Neural Networks

Predicting Slowdowns in Decadal Climate Warming Trends With Explainable Neural Networks

Controlled Abstention Neural Networks for Identifying Skillful Predictions for Regression Problems

Deep Learning for Multi‐Timescales Pacific Decadal Oscillation Forecasting

Contact Info

Product

Resources

About