Aggregation strategies to improve XAI for geoscience models that use correlated, high-dimensional rasters

Krell, Evan; Kamangir, Hamid; Collins, Waylon; King, Scott A.; Tissot, Philippe

doi:10.1017/eds.2023.39

Cited by 4 publications

(1 citation statement)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, where possible and appropriate, closely related variables may be grouped or transformed for collective or conditional interpretation, where their contributions can be considered more holistically, rather than attempting to separate the individual contributions of these variables (Brenning, 2023;Jiang et al, 2024). Krell et al (2023) further suggest that models based on gridded geospatial data can be sensitive to the choice of grouping scheme, and thus it is beneficial to compare explanations from multiple grouping schemes for more accurate insights, as each may probe the model differently.…”

Section: Multicollinearity and Dependence Among Featuresmentioning

confidence: 99%

How Interpretable Machine Learning Can Benefit Process Understanding in the Geosciences

Jiang,

Sweet,

Blougouras

et al. 2024

Earth's Future

View full text Add to dashboard Cite

Interpretable Machine Learning (IML) has rapidly advanced in recent years, offering new opportunities to improve our understanding of the complex Earth system. IML goes beyond conventional machine learning by not only making predictions but also seeking to elucidate the reasoning behind those predictions. The combination of predictive power and enhanced transparency makes IML a promising approach for uncovering relationships in data that may be overlooked by traditional analysis. Despite its potential, the broader implications for the field have yet to be fully appreciated. Meanwhile, the rapid proliferation of IML, still in its early stages, has been accompanied by instances of careless application. In response to these challenges, this paper focuses on how IML can effectively and appropriately aid geoscientists in advancing process understanding—areas that are often underexplored in more technical discussions of IML. Specifically, we identify pragmatic application scenarios for IML in typical geoscientific studies, such as quantifying relationships in specific contexts, generating hypotheses about potential mechanisms, and evaluating process‐based models. Moreover, we present a general and practical workflow for using IML to address specific research questions. In particular, we identify several critical and common pitfalls in the use of IML that can lead to misleading conclusions, and propose corresponding good practices. Our goal is to facilitate a broader, yet more careful and thoughtful integration of IML into Earth science research, positioning it as a valuable data science tool capable of enhancing our current understanding of the Earth system.

show abstract

Section: Multicollinearity and Dependence Among Featuresmentioning

confidence: 99%

How Interpretable Machine Learning Can Benefit Process Understanding in the Geosciences

Jiang,

Sweet,

Blougouras

et al. 2024

Earth's Future

View full text Add to dashboard Cite

show abstract

Predicting Cloud‐To‐Ground Lightning in the Western United States From the Large‐Scale Environment Using Explainable Neural Networks

Kalashnikov,

Davenport,

Labe

et al. 2024

JGR Atmospheres

View full text Add to dashboard Cite

Lightning is a major source of wildfire ignition in the western United States (WUS). We build and train convolutional neural networks (CNNs) to predict the occurrence of cloud‐to‐ground (CG) lightning across the WUS during June–September from the spatial patterns of seven large‐scale meteorological variables from reanalysis (1995–2022). Individually trained CNN models at each 1° × 1° grid cell (n = 285 CNNs) show high skill at predicting CG lightning days across the WUS (median AUC = 0.8) and perform best in parts of the interior Southwest where summertime CG lightning is most common. Further, interannual correlation between observed and predicted CG lightning days is high (median r = 0.87), demonstrating that locally trained CNNs realistically capture year‐to‐year variation in CG lightning activity across the WUS. We then use layer‐wise relevance propagation (LRP) to investigate the relevance of predictor variables to successful CG lightning prediction in each grid cell. Using maximum LRP values, our results show that two thermodynamic variables—ratio of surface moist static energy to free‐tropospheric saturation moist static energy, and the 700–500 hPa lapse rate—are the most relevant CG lightning predictors for 93%–96% of CNNs depending on the LRP variant used. As lightning is not directly simulated by global climate models, these CNNs could be used to parameterize CG lightning in climate models to assess changes in future CG lightning occurrence with projected climate change. Understanding changes in CG lightning risk and consequently lightning‐caused wildfire risk across the WUS could inform fire management, planning, and disaster preparedness.

show abstract

Interpretation and Attribution of Coastal Land Subsidence: An InSAR and Machine Learning Perspective

Qiao,

Chu,

Krell

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

Subsidence, the downward vertical land motion (VLM), plays a pivotal role in contributing to the risk of coastal flooding. Accurately estimating VLM and identifying its potential features related to subsidence can provide crucial information for stakeholders to make better informed decisions. This study aimed to estimate large-scale subsidence at the Texas Gulf Coast and identify potential subsidence features using explainable models. Nine potential features were considered for modeling the VLM, ranging from natural terrain variations to anthropogenic activities. These features were used to train a random forest (RF) machine learning (ML) model. Explainable artificial intelligence (XAI) techniques including SHapley Additive exPlanations (SHAP) and impurity-and permutation-based feature importance were used to identify the contributions to subsidence. The results demonstrated favorable performance of the RF model, achieving an R 2 value of 0.56 during validation. XAI results underscored the significance of the digital elevation model (DEM) on explaining subsidence at the Texas Coast. Additionally, XAI analysis highlighted the overall contribution of subsidence from anthropogenic activities, such as hydrocarbon extraction and groundwater withdrawal. Furthermore, the sample-level SHAP map provided detailed and reasonable subsidence-attribution results across the study area, showing potentials for automatic and data-driven explanations of the VLM.

show abstract

Aggregation strategies to improve XAI for geoscience models that use correlated, high-dimensional rasters

Cited by 4 publications

References 41 publications

How Interpretable Machine Learning Can Benefit Process Understanding in the Geosciences

How Interpretable Machine Learning Can Benefit Process Understanding in the Geosciences

Predicting Cloud‐To‐Ground Lightning in the Western United States From the Large‐Scale Environment Using Explainable Neural Networks

Interpretation and Attribution of Coastal Land Subsidence: An InSAR and Machine Learning Perspective

Contact Info

Product

Resources

About