How it works
How the watering advice is worked out, and how it was checked
The advice follows the FAO-56 daily water balance, the method used in irrigation planning worldwide. This page shows the data it uses, the checks run so far, and what is still untested.
The method in one picture
Each day the soil loses water to the crop (crop water use = reference evapotranspiration × a crop coefficient) and gains it from rain. When the soil has lost the share of its stored water the crop can easily spare, watering is due. The size of the water store depends on soil and rooting depth.
Water missing today = yesterday − effective rain + crop water use
Water is due when the shortage reaches the readily available water (RAW = p × TAW).
Where the data comes from
| Source | What it gives | Status |
|---|---|---|
| Open-Meteo | Forecast and recent weather (ECMWF and ERA5 based): reference evapotranspiration, rain | In use. CC BY 4.0. |
| FAO Paper 56 | The method: crop coefficients, stage lengths, rooting depth, depletion fraction | In use. The water balance follows it. |
| NASA POWER | Independent weather record back to 1981, used to check the forecast source | In use for checks. |
| SoilGrids (ISRIC) | Sand, silt and clay at 250 m, to suggest a soil class for a location | Being added; the service was unavailable when tested, so it falls back to a soil you choose. |
| India Meteorological Department | Gridded rainfall and temperature, official forecasts | Not in use. Reference for later checks. |
| ICAR and state agriculture universities | Soil and crop references for India | Reference for future field comparison. |
Weather data by Open-Meteo.com, CC BY 4.0.
What we have checked so far
Three checks, all done on a computer. None uses measurements from a real field yet.
1. The arithmetic
We wrote the water balance a second time independently and ran both on 14 crops and 3 soils over 80 days. Every day matched (0 mismatches). This shows the code does what the method says, not that the method fits a given field.
2. Two weather sources compared
The two sources disagree by roughly 1 mm of daily crop water use, most around Theni and Coimbatore. Over ten days that is about one watering, so the date can move by a few days. We plan to show a range, not a single date.
3. Against an independent FAO-56 program
In the middle and late season the two agree within about 3%. In the first month our simpler single-coefficient method uses more water than the dual-coefficient program, which makes the first watering come about ten days early in this test. We plan to correct this.
What has not been tested
- No soil-moisture readings from a real field have been compared with the advice.
- Soil values are typical for a soil class, not measured on your field.
- Rain at your field can differ from the forecast for your area, and forecasts weaken with distance in time.
- Crop values come from FAO tables for general conditions, not local trials.
Published irrigation tools are checked the same way: against a reference calculation, then against soil sensors and farmers' own practice over several seasons. For example, a forecast-based scheduling study at IIT Bombay on two Nashik farms reported 10–30% less irrigation than farmers' practice with no significant yield loss (Roy et al., Water Resources Research, 2021). Field comparison is our next step.
Questions or a field to test on? Contact MM Rooted Ventures. Research by MM Rooted Ventures, Theni and Bengaluru.