Validation dataset for an integrated simulation framework for heat and mass transfer in soilless, naturally ventilated greenhouses with row crop growth

Description

This dataset supports the validation results reported in the article "An integrated simulation framework for heat and mass transfer in soilless, naturally ventilated greenhouses with row crop growth". It contains the simulated and measured series for a commercial paprika greenhouse in the Republic of Korea over a 278-day cultivation cycle, from 14 February to 18 November 2023, at ten-minute resolution (40,032 records) together with hourly aggregates (6,672 records). Each simulated microclimate and control quantity is paired with the corresponding on-site record: Cropzone air temperature and relative humidity, the heating, ventilation, and vent-opening setpoints, and the thermal and shade screen positions. The external weather forcing and the simulated heating supply rate are also included; the latter is a model output for which no measured counterpart exists. A logging-gap flag and a ventilation-regime label are supplied so that the sample sizes reported in the article can be reproduced exactly (N = 39,962 ten-minute records and N = 6,656 hourly records). The deposit is scoped to the quantities that are compared against independent measurement; internal model intermediates that are not reported in the article are not included. The README documents the variable dictionary and the computation conventions required to reproduce the reported metrics, and verify.py recomputes every published evaluation metric directly from the two CSV files.

Steps to reproduce

1. Download the two CSV files, README.md, and verify.py into the same folder. 2. Install the dependencies: pandas and numpy. 3. Run: python verify.py The script reads HTVT_validation_10min.csv and HTVT_validation_hourly.csv, applies the logging-gap mask and the ventilation-regime label supplied in the files, and recomputes every evaluation metric reported in the article. Each reproduced value is printed next to the published value for direct comparison. The underlying simulation was performed with EnergyPlus v25.1 coupled to Python sub-models through the EnergyPlus Python EMS API, forced solely by the public Korea Meteorological Administration ASOS record. The measured series were logged by the on-site Priva climate controller and the in-zone sensor pair at the commercial greenhouse. The "Conventions required to reproduce the reported metrics" section of README.md documents the conventions required to reproduce the reported metrics, including the interval-end timestamp convention, the hourly aggregation rule, the Nash-Sutcliffe form of R-squared, and the screen binarization thresholds.

Categories

Funders

• Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry and Fisheries

  Gyeonggi-do, Anyang-si

  Grant ID: RS-2025-02221279
• National Research Foundation of Korea

  Ministry of Science and ICT

  Daejeon, Daejeon

  Grant ID: RS-2021-NR060130

Licence