SpaceCrop Habitat and Mars Crop VOI: exploratory crop tools, worked example and feedback request

Hi @PlantAWG colleagues,

I am Lissandra Kruse Fuganti, a master’s student in Applied Computing at UEPG, Brazil. I would like to share two exploratory crop research tools and invite feedback on datasets and validation. Ryan Scott suggested contacting the relevant AWG leads, and @dr.richard.barker Dr. Richard Barker’s feedback encouraged me to make the use case clearer with a worked example.

Project links and purpose

SpaceCrop Habitat: https://space-crop-adapt.fuganti.chatgpt.site/
Compare proposed temperature and humidity with records in a terrestrial crop table, using documented NASA OSD-766 and OSD-267 environmental profiles as reference scenarios. The interface also provides scientific details, additional exploratory simulators and an experiment-planning checklist.

Mars Crop VOI: https://mars-crop-voi.onrender.com/
Explore crop candidates through environmental resemblance alongside published terrestrial Mars-regolith-simulant evidence. The evidence-transfer probability framework is a hypothesis under explicit assumptions, not a validated estimate of cultivation success on Mars. The Render-hosted app may need time to wake up.

Try this one-minute example in SpaceCrop Habitat

  1. Open the SpaceCrop Habitat link and keep Orbit · space station selected.
  2. Click Reset initial values. Temperature is 23 °C and relative humidity 37.6%, from the OSD-766-SN001 profile. Kidney bean ranks first in this particular table comparison (90.8/100).
  3. Change only temperature to 25.5 °C, keeping humidity at 37.6%. Pigeon pea moves to first (94.6/100) and kidney bean to fifth (86.0/100). Results update immediately; no code or Run button is needed.
  4. Reset and change one input at a time to explore sensitivity. The optional pH comparison uses an illustrative starting pH of 6.5, not a measured pH from that NASA profile.

Screenshot: the 25.5 °C example

What this output means — and what remains untested

The practical question is: how does a shortlist of candidates change when a proposed environmental condition changes? Scores are environmental resemblance to the current terrestrial table, not probabilities of success, yield predictions or mission suitability recommendations. The table’s experimental provenance still requires verification. Light, CO2, airflow, substrate composition, cultivar, nutrients, crop size and duration, and spaceflight effects need further evaluation. A high-ranked perennial, for example, is not automatically a practical space crop.

For Mars, environmental resemblance and measured outcomes from terrestrial regolith-simulant experiments are distinct forms of evidence. Neither proves cultivation on Mars. The tools aim to help formulate hypotheses and plan ground experiments, not to claim a validated digital twin.

Feedback and publication work

I am preparing associated hypothesis manuscripts for publication. I would especially welcome suggestions on:

  • a suitable crop/cultivar, dataset and measured endpoint for an initial validation;
  • essential substrate and environmental variables currently missing;
  • appropriate comparators and independent experiments to test whether evidence transfer helps;
  • whether a Plant AWG or Regolith AWG subgroup would be a suitable place to develop this work.

Thank you for your guidance and constructive feedback.

Lissandra Kruse Fuganti — UEPG, Brazil

Prepared with AI assistance; shared as exploratory work for scientific discussion.

1 Like

Hi Plant AWG colleagues,

I have developed the soil-comparison idea into a reproducible analysis of public terrestrial measurements, focusing on Texas. This is a separate extension of my earlier crop-simulation work.

The USDA dataset contains 14,805 soil-layer records. I retained sample identifiers, depths, reported soil names, coordinates and analytical methods. I then compared relative Ca, Mg, Na and K profiles with individual Phoenix WCL reference samples using Aitchison distances. Chloride and sulfate were examined in additional sensitivity analyses; pH and electrical conductivity were kept as separate descriptive contrasts.

Among shallow horizons, the results identify Amarillo-named records in Hockley County and, under the conditional sulfate analysis, a Shippingpens-named record in Culberson County as leads for further investigation. The leading records change with the included ions and reference assumptions. They are not validated Mars analogues: perchlorate and ammonium are missing from the Texas panel, and saturation-extract measurements are not equivalent to Phoenix WCL extraction.

The earlier Brazilian screen found numerical pH overlaps in records from locations including Cruzeiro do Oeste, Guairaçá, Campo Mourão, Prudentópolis and Londrina. Those results do not establish broader chemical similarity or characterize all soils in those municipalities.

The revised Texas notebook ran successfully on Kaggle and generated 35 analysis files. The report explains the calculations, candidate records, uncertainty scenarios and measurements needed for a matched laboratory comparison.

GitHub repository
Texas notebook
Public Texas dataset
Brazil analysis code

Sources: Phoenix WCL results, Phoenix sulfate measurements, NASA PDS archive, USDA database, Census boundaries, and Brazilian soil dataset.

Best regards,
Lissandra Kruse Fuganti

Texas_soils_and_Mars_comparative_research.pdf (280.1 KB)