ABSTRACT:

This repository contains the machine-learning-ready datasets and Jupyter Notebook used in the study of groundwater storage and submarine groundwater discharge dynamics along the Arctic coast of Simpson Lagoon, Alaska. The related manuscript investigated the time-dependent influence of atmospheric, thermal, and oceanic controls on groundwater storage and discharge using three deep learning architectures (1D-CNN, LSTM, GRU).
The original groundwater and lagoon field observations collected during the 2022 thaw and summer seasons (June-September) are archived separately in Demir et al. (2026a). The datasets provided here consist of processed and quality-controlled groundwater lagoon observations merged with atmospheric and oceanic variables obtained from reanalysis products (MERRA-2 and GLDAS) and weather stations. Reanalysis atmospheric variables were obtained from NASA GES DISC (Acker & Leptoukh, 2007) a ~15 km × 15 km region surrounding the site. This repository also includes code for data pre-processing, deep learning model training, model evaluation, and eXplainable Artificial Intelligence (XAI) analyses using SHapley Additive exPlanations (SHAP). These resources enable full reproduction of the analyses used in the related manuscript, and provide a framework for applying XAI to Arctic coastal groundwater systems and other data-limited hydrologic systems.

References:
Acker, J. G., & Leptoukh, G. (2007). Online analysis enhances use of NASA earth science data. Eos, Transactions American Geophysical Union, 88(2), 14–17. https://doi.org/10.1029/2007EO020003
Demir, C., J. A. Guimond, E. Bristol, E. Bullock, J. W. McClelland, M. A. Charette, M. B. Cardenas (2026a). Hydrological and thermal field measurements at an Arctic coastal site, Simpson Lagoon, Alaska (2021-2022), HydroShare, https://doi.org/10.4211/hs.b2f0fcd9f2834f79a4b9216ad717eb69
Related manuscript:
Demir, C., Gomez-Velez, J. D., Guimond, J. A., McClelland, J. W., Charette, M. A., and Cardenas, M. B. Time-varying hydroclimatic and oceanic controls on Arctic submarine groundwater discharge inferred using explainable AI.

ABSTRACT:

Subterranean estuaries (STEs) along Arctic coasts are dynamic interfaces where supra-permafrost groundwater mixes with seawater and redistributes heat and solutes across the land-ocean continuum. This dataset documents seasonal groundwater and thermal dynamics in a supra-permafrost STE along a sandy beach at Simpson Lagoon, Alaska (Beaufort Sea coast), a low relief polygonal tundra landscape. Field observations were collected during three campaign-based deployments in 2022 (June 16-20, July 22-27, and September 29-October 5), representing early thawing, summer, and freeze-up conditions. Shore-perpendicular transects of PVC piezometers were installed repeatedly at key locations, including a recurring Transect-A and later-season Transect-E. In addition to the short-term, season-individual campaigns, Transect-A was continuously monitored from June to Oct, 2022. Screened section of the piezometers were restricted to the subsurface to measure phreatic surface elevations on the non-submerged beach and piezometric hydraulic head at the screen midpoint beneath lagoon-inundated sediments. A stainless-steel piezometer (PIce) was installed through sea ice into the subtidal seabed to monitor salinity variations and quantify lagoon-seabed vertical gradients under land-/bottom-fast ice. Groundwater and surface water levels, temperature, and electrical conductivity were monitored at 5-15 min intervals using Level-Temperature (LT) and Level-Temperature-Conductivity (LTC) pressure transducers (In-Situ and Solinst), with concurrent barometric logging (air temperature and pressure) and daily manual water-level measurements for calibration and reference elevation corrections. Conductivity was temperature-normalized to specific conductance at 25degC, and heads were converted to equivalent freshwater head to account for salinity effects. Thaw-depths along transects in three of the seasons were recorded using a steel probe. Subsurface temperature-depth profiles were measured using paired thermistor arrays (HOBO® 4-channel loggers and TrodX® multi-depth profilers) installed adjacent to selected piezometers during short-campaigns. In addition, thermistor arrays installed at the inter- to sub-tidal area monitored temperatures at multiple depths at 4-hr resolution over a year from August 2021 to July 2022. Site topography/instrument elevations were surveyed with a robotic total station (Trimble® S5) tied to GPS coordinates. All provided head and elevation data are based on a local reference elevation, however, likely monthly changes on the ground level due to thaw may have slightly influenced the reference elevation inter-seasonally. Field setup metadata, including coordinates, piezometer and thermistor design, and measurement elevations, was provided in this repository under "Field_setup_information" folder. All mentioned field observations can be found in the "Observations" folder.

This repository accompanies our manuscript: Demir, C., Guimond, J. A., Bristol, E. M., Bullock, E., McClelland, J. W., Charette, M. A., Cardenas, M. B. (In Review) Hydrological and thermal dynamics of a supra-permafrost subterranean estuary. This work used observation-constrained numerical simulations in COMSOL Multiphysics to resolve coupled groundwater flow and heat transport for three seasonal snapshot domains (early thawing, summer, early freeze-up) using campaign specific aquifer geometries (defined by ice-table extent) and initial/boundary conditions. Therefore, in addition to the processed field observations, this repository includes COMSOL Multiphysics model files (.mph) for June, July, and September domains, model outputs provided in the study (e.g. hydraulic head, temperature fields, and estimated water and heat fluxes), and Jupyter notebooks used to generate figures and analyses.