arXiv subjects

Michael J. Friedel

Publications and source records attributed to Michael J. Friedel.

4 recordsLinked to original sources

arXiv2026

Comparison of generative algorithms for conceptual groundwater modeling of coastal volcanic aquifer features with disparate, sparse and extremely imbalanced data

In coastal volcanic aquifers, the reliability of freshwater seawater-exchange simulations are governed by accuracy of the conceptual groundwater model (CGM). The traditional CGMs are constructed by qualitatively combining independent hydrogeophysical features, limiting their ability to capture the complexity of volcanic terrains. To integrate these disparate, sparse, and imbalanced features, we propose an AI-assisted workflow. First, the self-organizing map (SOM) is applied to estimate a deterministic set of transdisciplinary features called the reference model. Second, generative algorithms are applied to the reference model and empirical distributions constructed to obtain sets of stochastic point clouds called the site model. Data quality metrics identify the preferred generative algorithm whose set of stochastic features are mapped using SOM to the groundwater model grid and assigned as the stochastic CGM. The proposed algorithm is applied to extremely imbalanced multiclass features and multiple discrete numerical features observed at the Halawa-Moanalua aquifer, Oahu, Hawaii. At this stie, the Copula Generative Adversarial Network is deemed as the preferred generative algorithm whose set of stochastic transdisciplinary features represent the Halawa-Moanalua CGM. The simulated spatial geologic units correspond to published surface maps; and the simulated conductance, temperature, and barometric pressure profiles correlate with those measured at deep monitoring wells. Inspecting the 3-dimensional conductance models reveal groundwater flow and discharge driven by the aquifer hydraulic gradient, freshwater pumping, seawater intrusion induced by onshore withdrawals, and preferred pathways for freshwater-seawater exchange, such as landward intrusion of seawater and seaward discharge of freshwater.

physics.geo-ph↗

arXiv2024

A machine learning algorithm for predicting naturalized flow duration curves at human influenced sites and multiple catchment scales

Regional flow duration curves (FDCs) often reflect streamflow influenced by human activities. We propose a new machine learning algorithm to predict naturalized FDCs at human influenced sites and multiple catchment scales. Separate Meta models are developed to predict probable flow at discrete exceedance probabilities across catchments spanning multiple stream orders. Discrete exceedance flows reflect the stacking of k-fold cross-validated predictions from trained base ensemble machine learning models with and without hyperparameter tuning. The quality of individual base models reflects random stratified shuffling of spilt catchment records for training and testing. A Meta model is formed by retraining minimum variance base models that are bias corrected and used to predict final flows at selected percentiles that quantify uncertainty. Separate Meta models are developed and used to predict naturalised stochastic flows at other discrete exceedance probabilities along the duration curve. Efficacy of the new method is demonstrated for predicting naturalized stochastic FDCs at human influenced gauged catchments and ungauged stream reaches of unknown influences across Otago New Zealand. Important findings are twofold. First, independent observations of naturalised Median flows compare within few percent of the 50th percentile predictions from the FDC models. Second, the naturalised Meta models predict FDCs that outperform the calibrated SWAT model FDCs at gauge sites in the Taieri Freshwater Management Unit: Taieri at Tiroiti, Taieri at Sutton Creek, and Taieri River at Outram. Departures in the naturalised reference state are interpreted as flow regime changes across the duration curves. We believe these Meta models will be useful in predicting naturalised catchment FDCs across other New Zealand regions using physical catchment features available from the national data base.

physics.geo-ph↗

arXiv2023

Multimodal machine learning for 3-dimensional characterization of hidden groundwater and geothermal resources

We present a multimodal machine learning (MML) workflow to assimilate and simultaneously predict the 3d distribution of numeric and categorical features along a groundwater-geothermal continuum. Success of the MML workflow relies on a transductive learning algorithm that projects field modalities onto a single embedding space (hypersurface). Multimodalities can include any combination of measured (point field) and derived (multiphysics based numerical model inversions, data driven machine learning, and multiphysics informed machine learning) features. The proposed MML workflow is applied to assimilate randomly shuffled subsets of Hawaii Play Fairway modalities and predict subsurface geophysical, geologic, and hydrogeologic features at the Islands of Lanai and Hawaii. Despite challenging field data characteristics (disparate, scale dependent, spatially limited, sparse, and uncertain), the MML workflow yields a single 3d transdisciplinary model that generalizes well to independent data presented to the trained model. The predicted features are used to identify hidden groundwater and geothermal resources at Lanai, and geothermal resources at Hawaii. Other interpreted subsurface features at Lanai include basalt, batholith, dike swarm, pluton, sill, mantle, Moho, and 3d geothermal stratigraphic units; whereas interpreted subsurface features at Hawaii include 3d velocity layering, 3d earthquake-fault associations, 3d fault systems; basalt, oceanic crust, magmatic underplating, lithospheric flexure, mantle, and Moho. This study provides new capabilities for characterizing continuous subsurface groundwater and geothermal features for sustainable living in the Hawaiian Islands and other geothermal sites worldwide. Keywords: Machine learning, 3d hidden groundwater resources, hidden geothermal resources fault system, oceanic crust, magmatic underplating, lithospheric flexure, Moho, Hawaii

physics.geo-ph↗

arXiv2023

A simple stacked ensemble machine learning model to predict naturalized catchment hydrology and allocation status

New Zealand legislation requires that Regional Councils set limits for water resource usage to manage the effects of abstractions in over-allocated catchments. We propose a simple stacked ensemble machine learning model to predict the probable naturalized hydrology and allocation status across 317 anthropogenically stressed gauged catchments and across 18,612 ungauged river reaches in Otago. The training and testing of ensemble machine learning models provides unbiased results characterized as very good (R2 > 0.8) to extremely good (R2 > 0.9) when predicting naturalized mean annual low flow and Mean flow. Statistical 5-fold stacking identifies varying levels of risk for managing water-resource sustainability in over-allocated catchments; for example, at the respective 5th, 25th, 50th, 75th, and 95th percentiles the number of overallocated catchments are 73, 57, 44, 23, and 22. The proposed model can be applied to inform sustainable stream management in other regional catchments across New Zealand and worldwide.

physics.geo-ph↗