Crops and greenhouses
Greenhouse and open-field crops are where most of our work sits. The through line is the same in all of it: light, water and nutrients come in, the plant decides where to put them, and that decision shows up weeks later as kilos, quality, cost and water use.
Source and sink
The core crop model treats the plant as a supply of assimilates from photosynthesis, and a set of competing demands: leaves, stems, roots, and fruit at several stages of development. Yield is what falls out of that competition over time, not a growth curve fitted to it.
That framing is what lets the model answer steering questions. Why does a heavy fruit load now cause a gap five weeks later. What a warm night actually costs. When a planting date change pays and when it just moves the problem. Where the plant is limited by light and where it is limited by its own capacity to use it.
On top of the biology sit the layers that make it a business question: energy and water use, labour hours, price and contract structure, and the KPIs a crop manager reports on.
Try it: cropsimulator.com is a public, simplified version. Move the setpoints, watch the crop and the margin respond. No login.
Leaf stress
A shorter time base, sitting underneath the crop model. Energy and water balance at leaf level through a day: radiation load, transpiration, vapour pressure deficit, stomatal response, leaf temperature.
This is the layer where screens, cooling, misting and ventilation actually act, and where “the crop looked stressed at two in the afternoon” becomes a number. Because it runs sub-daily while the crop model runs on a longer clock, the two are coupled rather than merged, which keeps both honest.
Salinity, water reuse and drainage
Water is getting scarcer and more regulated, and water uptake makes salt accumulate. The salinity model tracks the salt balance of the root zone against irrigation quality, leaching fraction and blend ratio.
It is a small model with an unusually clear message, because the trade-off is real and two-sided: more salt costs yield and buys quality, and the water you save has to go somewhere. It also produces results that reverse standard advice, for example when desalination is available, where the leaching strategy that is right without it becomes wrong with it.
What a project here usually looks like
A grower or supplier has a decision with a long lag: a setpoint strategy, an irrigation regime, a planting schedule, an investment in equipment. The experiment to settle it would take two seasons and would only answer it for one season’s weather.
We build the model, calibrate what your data can support, say clearly what it cannot, and give you something you can run the alternatives on before you commit.