Mark S. Pleasants 2025 groundwater model Reston, VA, USA U.S. Geological Survey https://doi.org/10.5066/P1HH3X5W Mark S. Pleasants Michael N. Fienen Hedeff I. Essaid Joel D. Blomquist Jing Yang Ming Ye 2025 publication Water Resources Research vol. 61 n/a American Geophysical Union (AGU) https://doi.org/10.1029/2025WR040323 A synthetic one-dimensional groundwater flow model was developed using MODFLOW 6 as a test application of the prior-based model discrimination method described in the associated publication (https://doi.org/10.1029/2025WR040323). The model represents a steady-state unconfined aquifer where water flows horizontally between two constant head boundaries (a lake at the lefthand side and river at the righthand side) and precipitation recharges the aquifer. Process module uncertainty is considered in the representations of groundwater recharge, geologic structure, and snowmelt that controls the river stage constant head boundary. Two process modules are considered to represent each of the three uncertain system processes. This yielded eight candidate configurations of the groundwater model. This USGS data release contains all of the input and output files and ancillary Python scripts for the simulations and results described in the associated publication (https://doi.org/10.1029/2025WR040323). The synthetic models in this archive were developed as a test application of the prior-based model discrimination method described in the associated publication (https://doi.org/10.1029/2025WR040323). A network of synthetic data are created using a version of the 1-D groundwater model that is distinct from the eight candidate model configurations (referred to as the "true" model). These synthetic data are used during prior-based model discrimination to determine more-adequate model configurations prior to model calibration (model calibration is not performed here). A new metric, Mahalanobis distance deviation (M_D), is calculated to quantify the similarity of distributions of prior model outputs to the distributions of synthetic data. A new metric of data worth termed "discernment value" is then calculated to quantify the value of the synthetic data network for discerning more-adequate from less-adequate models prior to calibration. The development of the model input and output files included in this data release are documented in the associated publication (https://doi.org/10.1029/2025WR040323). Support is provided for correcting errors in the data release and clarification of the modeling conducted by the U.S. Geological Survey. Users are encouraged to review the model documentation report (https://doi.org/10.1029/2025WR040323) to understand the purpose, construction, and limitations of this model. The model will run successfully only if the original directory structure is correctly restored. The model archive is broken into several pieces to reduce the likelihood of download timeouts. Instructions for reconstructing the original directory structure and running the model included in this data release and described in the model documentation report can be found in the README.txt ASCII file which can be downloaded as part of this data release. All data and models in this archive are synthetic. These models and data are not based on a real location or time period, nor were the models calibrated (see associated publication for details: https://doi.org/10.1029/2025WR040323). For this reason, the time period information contained in this metadata file ("timeinfo") is listed as "Unknown" and the spatial domain bounding coordinates in this metadata file ("spdom") are listed as "-180.000, 180.0000, -90.0000, 90.0000". Please refer to the "Thumbnail.jpg" file in this archive for information on the model design. Files in this data release include: -README.txt: ASCII text file that describes the model data release. This file also includes instructions on how to run the models contained in this data release. -Thumbnail.jpg: JPEG image showing the synthetic model domains for the candidate and true models. -batch.zip: Zipped directory containing a csv file that defines the parameter realizations that are to be run using the candidate models. -environment.zip: Zipped directory containing an environment.yml file that can be used to create a conda environment to run the Python scripts contained in this archive. -mf6_htc_sims.zip: Zipped directory containing MODFLOW 6 input and output files of the candidate models. -mf6_true_sims.zip: Zipped directory containing MODFLOW 6 input and output files of the synthetic "true" model. -reals.zip: Zipped directory containing model parameter realizations. -results.zip: Zipped directory containing all model inputs/outputs and post-processed results. -scripts.zip: Zipped directory containing Python and bash scripts to run the workflow. Unknown See Supplemental Info Complete Not planned -180.0000 180.0000 90.0000 -90.0000 ISO 19115 Topic Category geoscientificInformation USGS Thesaurus groundwater modeling none usgsgroundwatermodel MODFLOW 6 Groundwater model Model discrimination Prior-data conflict Model structural uncertainty USGS Metadata Identifier USGS:67b8eb45d34e1a2e835b8476 None. Acknowledgement of the U.S. Geological Survey would be appreciated in products derived from this data release. These groundwater model input and output files and ancillary Python scripts are provided to support the analyses documented in the associated publication (https://doi.org/10.1029/2025WR040323). Although the information contained in this model archive may be useful for other purposes, it is incumbent on the user to understand the purpose, construction, and limitations of these models. Data have been checked to ensure consistency with the accompanying report. If any errors are detected, please notify the originating office. Mark Pleasants U.S. Geological Survey mailing and physical

521 Progress Circle, Suite 6

Cheyenne WY 82007 USA 1-307-778-2931 mpleasants@usgs.gov https://www.sciencebase.gov/catalog/file/get/67b8eb45d34e1a2e835b8476?name=Thumbnail.jpg Image of the synthetic model domains. (a) Groundwater model domain used to define the eight candidate models and (b) groundwater model domain used to create the synthetic data, i.e., the true model. jpg U.S. Geological Survey Water Availability and Use Science Program. None Unclassified None -All model simulations were run on an Amazon Web Services (AWS) cloud computing cluster with HTCondor. The mf6 files in this archive are the MODFLOW 6.5.0 version for Linux released on May 23, 2024 in the Github commit 879f911 (https://github.com/MODFLOW-USGS/modflow6/releases/tag/6.5.0). -The code was also tested on a personal computer with a 64-bit Windows 11 OS in the Windows Subsystem for Linux 2 (WSL2) after creating the environment given in the environment.yml file in the /environment folder. Using the parameter ensemble included in the /reals folder requires 6,000,000 simulations of the candidate MODFLOW 6 models. On the tested personal computer, this required approximately 12 hours of wall time running the models using 8 batches in parallel. To run the workflow using a smaller parameter ensemble, the variable N = 500 in params.py can be reduced. The total number of candidate model simulations is determined from the equation 3 * 8 * N * N. -Using the synthetic data set size of 500 given by the variable N = 500 in sims_true.py, the true model was simulated 250,000 times. On the tested personal computer, this required approximately 3 hours of wall time running all simulations in serial. To run the workflow using a smaller synthetic data set size, the variable N = 500 in sims_true.py can be reduced. The total number of true model simulations is determined from the equation N * N. Christian D. Langevin Joseph D. Hughes Edward R. Banta Richard G. Niswonger Sorab Panday Alden M. Provost 2017 publication Techniques and Methods 6-A55 Reston, VA, USA US Geological Survey Detailed description of the model input and output files included in this data release can be found in this model code documentation report. https://doi.org/10.3133/tm6A55 Jing Yang Ming Ye Xingyuan Chen Heng Dai Anthony P. Walker 20220302 publication Water Resources Research vol. 58, issue 3 n/a American Geophysical Union (AGU) The MODFLOW 6 model is based on the analytical groundwater model used in this reference. https://doi.org/10.1029/2021WR029812 All data in this archive are synthetic and models were not calibrated. No formal logical accuracy tests were conducted. Data set is considered complete for the information presented, as described in the abstract. Users are advised to read the rest of the metadata record and the associated model documentation report (https://doi.org/10.1029/2025WR040323) for additional details. No formal positional accuracy tests were conducted. No formal positional accuracy tests were conducted. The process used to develop and apply the model is fully described in the associated publication (https://doi.org/10.1029/2025WR040323). 2025 This model archive data release contains all the model input and output files needed to replicate the simulations described in the associated publication (https://doi.org/10.1029/2025WR040323). https://doi.org/10.1029/2025WR040323 Detailed descriptions of the model input and output files can be found in the MODFLOW 6 model documentation (https://doi.org/10.3133/tm6A55) and online at: https://water.usgs.gov/water-resources/software/MODFLOW-6/mf6io_6.5.0.pdf accessed February 24, 2025. https://doi.org/10.3133/tm6A55 GS ScienceBase U.S. Geological Survey mailing and physical

Denver Federal Center, Building 810, Mail Stop 302

Denver CO 80225 United States 1-888-275-8747 sciencebase@usgs.gov Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although the data, software, and related material have been processed successfully on a computer system at the U.S. Geological Survey (USGS), reviewed for accuracy and completeness, and approved for release by the USGS, no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty. Although the data have been subjected to rigorous review and are substantially complete, the USGS reserves the right to revise the data pursuant to further analysis and review. Furthermore, the data are released on the condition that neither the USGS nor the U.S. Government shall be held liable for any damages resulting from authorized or unauthorized use. The USGS or the U.S. Government shall not be held liable for improper or incorrect use of the data described and/or contained herein. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Digital datasets None 4310.194 https://doi.org/10.5066/P1HH3X5W None. No fees are applicable for obtaining the dataset. 20250903 Mark S. Pleasants U.S. Geological Survey mailing and physical

521 Progress Circle, Suite 6

Cheyenne WY 82007 USA 1-307-778-2931 mpleasants@usgs.gov Content Standard for Digital Geospatial Metadata FGDC-STD-001-1998