Mendeley Data Showcase

# Biscayne Aquifer POMDP Experiment Companion code for: > Jakaite, L. & Schetinin, V. (2026). *A Unified POMDP Framework with Robust Satisficing, Directed Learning, and Real Options for Environmental Decisions under Deep Uncertainty.* Environmental Modelling & Software (submitted). --- ## Repository structure ``` biscayne/ biscayne_main.py # Main experiment — runs all 6 strategies make_graphical_abstract.py # Generates graphical abstract figure modules/ data_loader.py # USGS/NOAA groundwater data loader enso_loader.py # NOAA CPC ONI index loader regime_calibrator.py # HMM climate regime calibration (ERA5) obs_calibrator.py # Observation noise calibration salinity_proxy.py # Physics-based salinity proxy data/ raw/ # ERA5, NOAA sea-level, ONI, salinity proxy CSVs calibration_params.json # Pre-computed HMM parameters transition_matrix.csv # Calibrated regime transition matrix results/ dapp_final/ # Canonical results cited in the paper (n=1,000) mc_results.json # 6-strategy failure rates and costs fig_mc_comparison.png fig_barrier_timing.png fig_belief_trajectory.png fig_sensitivity_heatmap.png concept_colab.py # Conceptual illustration of Algorithm 1 requirements.txt ``` --- ## Reproducing the results ```bash pip install -r requirements.txt cd biscayne python biscayne_main.py --mc-trials 1000 --seed 42 ``` Expected output (`results/mc_results.json`): | Strategy | Failure rate | Cost | |---|---|---| | Unified Framework | 0.0 % | 34.2 | | RDM-Conservative (Lempert 2006) | 0.0 % | 35.7 | | DAPP (Haasnoot 2013) | 9.5 % | 118.9 | | Myopic POMDP | 41.6 % | — | | Robust Satisficing | 43.4 % | — | | BDT Baseline | 59.1 % | 393.9 | Runtime: ~3–5 minutes on a standard laptop (Python 3.10+). --- ## Data sources | Dataset | Source | Period | |---|---|---| | ERA5 precipitation | ECMWF open data | 2000–2023 | | NOAA Key West sea level | NOAA Tides & Currents | 2000–2023 | | ENSO ONI index | NOAA CPC | 2000–2023 | | Physics-based salinity proxy | Computed (see `modules/salinity_proxy.py`) | 2000–2023 | --- ## Conceptual illustration (optional) `concept_colab.py` is an interactive illustration of Algorithm 1 on a synthetic 6-step trajectory. It does not reproduce the paper's quantitative results but demonstrates the belief-update and decision logic in isolation. ```bash python concept_colab.py # Outputs: output/f4_reproduced.pdf, output/f5_reproduced.pdf ``` --- ## License MIT — see `LICENSE`.

This depository is the result of a calibrated microsimulation that generated the first individual-level dataset for Ghana’s Basic National Pension Scheme (Tier 1) for the period 2015 to 2024. The dataset combines publicly available annual aggregates from the Social Security and National Insurance Trust (SSNIT), National Pensions Regulatory Authority (NPRA) and the Ministry of Labour, Jobs and Employment (MLJE) with demographic, labour and mortality inputs from the Ghana Statistical Service (GSS) and the United Nations World Population Prospects. Using stratified sampling, membership transition modeling and poststratification calibration, the dataset recreates contributors, retirees and non-members at fine age-sex resolution while exactly reproducing official totals on contributions and benefits/pensions. The resulting microdata support distributional analysis of earnings, contributions, pensions and replacement rates, as well as actuarial projections, policy simulations and research on pension adequacy, coverage expansion and long-term social protection financing in Ghana.

This data is for the U.S. Corn Belt (US-CB), a major agricultural region characterized by intensive corn cultivation. The data includes 210 counties consistently dominated by corn production, selected based on their suitability for sub-pixel SIF disaggregation and inclusion in the USDA’s annual yield reporting. The data include five growing seasons: 2015, 2016, 2018, 2019, and 2020. The year 2017 was excluded from the data due to incomplete satellite data caused by sensor malfunction. Annual corn yield records were obtained from the U.S. Department of Agriculture’s National Agricultural Statistics Service (USDA-NASS). County-level yield data, measured in tons per hectare, are generated using a combination of farmer-reported surveys, field assessments, and calibrated simulation models. This integration ensures their reliability as ground-truth references for both training and validating predictive models. Their consistent spatial resolution and broad temporal span across the U.S. Corn Belt make them a solid basis for model development and performance testing. Standard satellite-derived SIF products are provided at coarse spatial resolutions, complicating their use in heterogeneous agricultural landscapes. To overcome this limitation, I applied a sub-pixel extraction method developed by Kira and Sun (2020), which uses high-resolution land cover and crop maps to isolate the contribution of corn within each SIF pixel. This technique allowed for the derivation of corn-specific SIF values at the county level, minimizing contamination from other crops or land covers.

This dataset supports the manuscript “Renewable curtailment under China’s mechanism-settled electricity reform: A three-party evolutionary game analysis of green certificate pricing”, submitted to Energy Policy. The dataset contains the processed inputs and simulation materials used to examine whether green electricity certificate (GEC) prices can contribute to renewable curtailment reduction after China’s Notice No. 136. It includes author-compiled renewable utilization data for Heilongjiang Province in 2025, official monthly GEC price anchors, policy-stage classification, scenario-calibrated model parameters, GEC-price perturbation settings, and simulation outputs for the August transition-stage and October post-implementation experiments. The renewable utilization data are derived from publicly available official and sectoral sources and are processed into monthly wind, photovoltaic, and composite renewable utilization indicators. The GEC price anchors are based on official monthly green certificate market bulletins. Other model parameters, including mechanism-settlement proxies, market-price proxies, behavioral parameters, and sensitivity ranges, are scenario-calibrated values used for mechanism-testing simulation rather than directly observed market data. The dataset is intended to improve transparency and reproducibility of the perturbation experiments reported in the manuscript. The accompanying simulation code or pseudocode reproduces the main scenario-calibrated results, including the identification of stage-specific effective GEC-price windows and the comparison between August and October policy environments.

This dataset contains all raw and processed experimental data, SEM images, and finite element simulation files supporting the findings of the related research article. Includes DMA data, shape memory test results, recovery stress measurements, calibrated Prony series parameters, WLF constants, Abaqus input files, and validation plots.

Calibrated dataset created for set-theoretic analysis of the relations between Democracy, Rule of Law, Internationalisation, Corporate Governance, and GNIpc in producing the outcome of CSR adoption.

This file contains simulation model input files for the simulations carried out including a windows executable and the source code for the simulation software OGS v5. details: - exe/ windows executables for the software ogs5 - ogs5/ source code of ogs5 - calibration/ recalibrated process models, model calibrated on data calibration by Boog (2019) - validation/ validated process model, model validated on data cross-validation by Boog (2019) - scenario_analysis/ process model used for scenario_analysis - how_to_run_simulations.pdf/ guideline how to run simulations on win and linux OS - file_orga.csv details of model input files

This package accompanies the manuscript on an integrated SysML–Modelica workflow for propulsion-system modeling, co-simulation, and validation. The purpose is to: (1) demonstrate the end-to-end workflow; and (2) allow readers to execute the supplied Modelica/co-simulation project and inspect the system structure and interface mapping. The model framework and workflow artifacts are shared in full. The parameter sets and datasets included in this package are synthetic or modified such that the resulting simulation outputs resemble the trends shown in the paper, but they do not correspond to the confidential bench-calibrated values used in the actual experiments. They are provided solely to verify workflow consistency and to allow users to experience the complete simulation and validation process. These files therefore enable reproduction of the modeling and validation workflow, but not reproduction of the quantitative experimental results. Users wishing to obtain quantitative fidelity on their own hardware should perform parameter identification/calibration using their own experimental data and controller–actuator configuration.

This dataset supports the research paper titled “Gulf Shock Transmission and Macroeconomic Fragility in Developing Economies: A Bayesian-Calibrated Monte Carlo Vulnerability Assessment.” The study examines how a Gulf-centred geopolitical shock may transmit to selected developing economies through GCC remittance exposure, energy-import pressure, shipping and logistics disruption, external-demand compression, foreign-exchange stress, inflation, reserve loss, policy-rate response, growth vulnerability and governance-fiscal fragility. The dataset covers six economies: Pakistan, Bangladesh, Sri Lanka, India, the Philippines and Bhutan. It includes raw remittance source files, processed remittance exposure files, macroeconomic and structural exposure inputs, recursive simulation outputs, Bayesian-calibrated Monte Carlo results, final tables, figures, audit files and R scripts used to reproduce the analysis. GCC remittance exposure is calculated as GCC-origin remittances divided by total recipient-country remittances, using Saudi Arabia, United Arab Emirates, Kuwait, Qatar, Oman and Bahrain as GCC source economies. The research develops a recursive Macroeconomic Vulnerability Index (MVI) and applies symmetric and asymmetric Gulf-shock scenarios. The Bayesian-calibrated Monte Carlo layer uses 10,000 draws to generate median vulnerability estimates, 90% credible intervals, uncertainty widths and rank-probability matrices. The results show a stable but intensity-sensitive vulnerability hierarchy, with Pakistan and Sri Lanka emerging as the highest-vulnerability cases under the extreme-combined scenario, followed by Bangladesh, India, the Philippines and Bhutan. The dataset is organised to allow replication of the paper’s tables, figures and Monte Carlo results. The R scripts can be run from the IRANPAK project folder to reproduce the cleaned remittance architecture, scenario-simulation outputs, Bayesian uncertainty tables, figures and audit logs. The files are provided for transparency, verification and future extension of the study, including possible full-panel macroeconomic modelling, alternative governance-fiscal amplification structures and robustness analysis. Data sources include public central-bank remittance publications, World Bank remittance and macroeconomic data, IMF and UNCTAD policy sources, and processed scenario-simulation files generated by the authors. Users should cite the original public data sources where relevant and cite this dataset when using the processed replication files, simulation outputs or R scripts.

Daily precipitation, runoff, and sediment yield data in a 'deadwood + grass' covered plot in the Eastern Italian Alps, together with climate, soil, slope, and plant/management input files for uncalibrated, partially calibrated, and fully calibrated WEPP simulations.