Transparency
Methods and data
What the app measures, how scores are built, and what it cannot tell you.
Current data source
Source: Demonstration data (not NASA observations)
Units: ndvi: index 0-1 · soilMoisture: m³/m³ · landSurfaceTemp: °C · precipitation: mm/month
Limitations: Illustrative values only. Real products differ in resolution, revisit time and quality.
The app never presents demonstration values as NASA observations. The badge and source line change automatically when a valid NASA file is loaded.
How the overall score is built
Overall score = Σ (dimension score × weight), rounded to 0–100. Weights are provisional design values and are re-normalised after priorities and season setup are applied.
- Soil benefit
- Legume nitrogen fixation, a following nitrogen-demanding crop, same-family penalties, and soil-type fit.
- Water fit
- Average crop water demand compared with an estimated supply for rain-fed, sprinkler or drip irrigation.
- Climate fit
- Heat tolerance against the observed heat-risk value.
- Diversity
- The number of distinct crop families in the sequence.
- Season fit
- How close each crop is to your chosen sowing month and cool/warm preference. It is not scored when neither is set.
- Priority alignment
- The average of the dimensions your selected priorities map to. It is not scored when no priorities are chosen.
What the observation layers mean
NDVI
A vegetation greenness index. Useful for seasonal vigor and anomalies; not a direct measure of yield or nutrients.
Soil moisture
Satellite products such as SMAP estimate near-surface moisture at coarse resolution. Confirm with a probe or hand test.
Land surface temperature
The temperature of the surface, not the air. Useful for heat-stress screening, especially around flowering.
Precipitation
Gridded rainfall products such as GPM IMERG can differ from a field rain gauge, especially in hilly terrain.
Adding NASA data when it arrives
Place a file named data/nasa_observations.json on the server. The app validates that every series has the same length, then switches the badge and source line automatically. If the file is missing or invalid, demonstration data stays active.
{
"meta": {
"source": "NASA product name and version",
"updated": "YYYY-MM-DD",
"units": {
"ndvi": "index 0-1",
"soilMoisture": "m³/m³",
"landSurfaceTemp": "°C",
"precipitation": "mm/month"
},
"limitations": "Resolution, revisit time and quality notes",
"products": ["MODIS", "SMAP", "ECOSTRESS", "GPM IMERG"]
},
"labels": ["Jan", "Feb", "Mar", "Apr", "May", "Jun",
"Jul", "Aug", "Sep", "Oct", "Nov", "Dec"],
"ndvi": [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
"soilMoisture": [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
"landSurfaceTemp": [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
"precipitation": [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
"heatRisk": 2.5
}
The example uses zero placeholders only to show the shape. Do not upload it as real data. Replace the zeros with the received observations and document the product, units, dates, resolution and limitations in meta. All-zero series are rejected so placeholders cannot be shown as NASA observations.
Crop parameters in the model
| Crop | Family | Season | Water | N demand | N fixation | Heat tolerance | Sowing months |
|---|---|---|---|---|---|---|---|
| Wheat | Grass | Cool | 2/3 | 3/3 | 0/3 | 2/3 | March, October, November |
| Barley | Grass | Cool | 1/3 | 2/3 | 0/3 | 2/3 | March, April, October, November |
| Maize | Grass | Warm | 3/3 | 3/3 | 0/3 | 2/3 | April, May, June |
| Sorghum | Grass | Warm | 1/3 | 2/3 | 0/3 | 3/3 | May, June |
| Soybean | Legume | Warm | 2/3 | 1/3 | 3/3 | 2/3 | May, June |
| Lentil | Legume | Cool | 1/3 | 1/3 | 2/3 | 2/3 | March, April, October, November |
| Canola | Brassica | Cool | 2/3 | 2/3 | 0/3 | 1/3 | March, April, September, October |
| Sunflower | Asteraceae | Warm | 2/3 | 2/3 | 0/3 | 3/3 | April, May |
| Potato | Nightshade | Cool | 3/3 | 3/3 | 0/3 | 1/3 | March, April, May |
These are illustrative design values. Replace them with sourced FAO or national extension guidance before using the app for real decisions.
Honest limits and privacy
- Crop parameters and weights are illustrative design values, not validated agronomy.
- Scores are relative model estimates, not predictions of yield, profit or soil nutrient levels.
- Missing information is never scored as favourable. Costs and revenue are intentionally not estimated.
- The Field Explorer is a labelled schematic, not satellite imagery. The included field picture is an illustration.
- Farm setup, history and plans are stored in the PHP session and in data/state.json on your server. There is no account, tracking script or third-party service.