> ## Documentation Index
> Fetch the complete documentation index at: https://badixth-dc85e378.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# Healthy NDVI Ranges for Paddy Across Rice Ecosystems
> Indicative healthy NDVI bands for paddy across irrigated lowland, rainfed lowland, rainfed upland, and flood-prone ecosystems, by growth phase.
Healthy NDVI for paddy is not a single curve. The same variety planted in an irrigated lowland versus a rainfed upland field will read differently on satellite imagery because of water background, canopy density, and establishment method. This page lists indicative healthy NDVI bands by rice ecosystem so you can interpret the Health tab correctly for the field in front of you.
For the base paddy preset and the generic Health tab workflow, see [Crop Health](/guides/crop-health-monitoring).
## Rice ecosystems at a glance
| Ecosystem | Water regime | Establishment |
| ----------------------- | ----------------------------------------- | ----------------------------- |
| Irrigated lowland | Controlled flooding, bunded | Transplanted (mostly) |
| Rainfed lowland | Bunded, rainfall-dependent standing water | Transplanted or direct-seeded |
| Flood-prone / deepwater | Prolonged deep flooding (> 50 cm) | Transplanted or broadcast |
| Rainfed upland | Aerobic soil, no standing water | Direct-seeded |
There is no single peak NDVI for an ecosystem. A band resolves against ecosystem **and** growth phase together, in the per-ecosystem tables below. See [How a band resolves](/snippets/band-resolution).
Ranges on this page are indicative bands drawn from published Sentinel-2 and Landsat rice studies and IRRI agronomy references. The platform refines them automatically once it has one or more seasons of field history.
## A reading above the band
The band describes what healthy looks like at this phase in this ecosystem. It is not a floor. **A reading above the band is a finding, not the absence of one** — see [How a band resolves](/snippets/band-resolution).
Paddy is the crop where this matters most, because paddy is the crop whose canopy is **supposed to fall**.
**Through ripening and maturity, high means late, not healthy.** Every ecosystem table on this page drops sharply at the last phase — irrigated lowland goes from 0.75 – 0.90 at heading to 0.40 – 0.65 at ripening. That decline is the crop moving nitrogen and sugars out of the leaves and into the grain. A field still reading heading values at day 110 has **not** got a better canopy than its neighbours; it has delayed maturity. The usual causes are late or excess nitrogen, a variety longer than the one on record, or a late establishment date. All three matter, because all three move the harvest date and two of them raise lodging risk.
**Through the first 30 days, high usually means weeds.** A field that reaches high NDVI too early is often weedy rather than vigorous — the weeds are contributing greenness while the rice stand underneath is poor. This is documented in more detail under [grass weeds](/guides/Crop/Rice/biotic/weeds/grasses).
**Through heading and flowering, the question cannot be asked.**
**Saturation carve-out.** NDVI saturates above \~0.8 in every rice ecosystem, which is inside the heading and flowering band on irrigated and rainfed lowland. The platform raises no above-band finding there, and says the reading has saturated before naming its band.
**Rainfed upland is the exception.** It rarely saturates, so an above-band reading stays readable across the whole season on upland fields.
Ripening is the one phase in this documentation where an above-band reading is both readable and consequential on every ecosystem. If only one above-band rule is ever implemented, it should be this one.
## Irrigated lowland
Controlled water depth, uniform stand, and dense transplanted canopy. This is the baseline used by the default paddy preset in [Crop Health](/guides/crop-health-monitoring).
| Phase | Days after transplanting | Healthy NDVI | Notes |
| -------------------------------- | ------------------------ | ------------ | ------------------------------------------------ |
| Transplanting / early vegetative | 0 – 20 | 0.20 – 0.40 | Water visible between hills depresses NDVI. |
| Tillering | 20 – 45 | 0.40 – 0.65 | Canopy closes evenly under stable water. |
| Panicle initiation / booting | 45 – 70 | 0.65 – 0.80 | Steady rise, low field-to-field variance. |
| Heading / flowering | 70 – 95 | 0.75 – 0.90 | Peak NDVI, saturates. Use NDRE for finer detail. |
| Ripening / maturity | 95 – 120 | 0.40 – 0.65 | Decline expected as grain fills. |
## Rainfed lowland
Bunded fields that rely on rainfall to maintain standing water. Water depth fluctuates, and short dry spells during tillering are common. Expect slightly lower peaks and a noisier time series than irrigated lowland.
| Phase | Days after establishment | Healthy NDVI | Notes |
| ---------------------------- | ------------------------ | ------------ | ------------------------------------------------------------ |
| Early vegetative | 0 – 25 | 0.20 – 0.40 | Add \~5 – 10 days if direct-seeded rather than transplanted. |
| Tillering | 25 – 50 | 0.35 – 0.60 | Uneven tillering from dry spells is normal. |
| Panicle initiation / booting | 50 – 75 | 0.60 – 0.80 | Watch for stalled NDVI rise, a sign of moisture stress. |
| Heading / flowering | 75 – 100 | 0.70 – 0.85 | Peak slightly below irrigated lowland. |
| Ripening / maturity | 100 – 125 | 0.40 – 0.65 | Decline may start earlier if end-of-season rains fail. |
On rainfed lowland fields, prioritize the **Anomaly** view over absolute NDVI once you have one season of history. It absorbs weather-driven variability that the absolute band cannot.
## Flood-prone / deepwater
Fields subject to prolonged deep flooding, typically in river deltas and low-lying coastal areas. Standing water and elongated plant architecture depress NDVI throughout the season, and flood pulses cause temporary dips even in healthy crops.
| Phase | Days after establishment | Healthy NDVI | Notes |
| ---------------------- | ------------------------ | ------------ | ----------------------------------------------------------------- |
| Early vegetative | 0 – 30 | 0.15 – 0.35 | Deep water background lowers early NDVI. |
| Tillering / elongation | 30 – 60 | 0.30 – 0.55 | Plants elongate rather than tiller heavily; canopy stays sparser. |
| Reproductive | 60 – 90 | 0.50 – 0.75 | NDVI recovers as canopy rises above water. |
| Heading / flowering | 90 – 115 | 0.55 – 0.80 | Peak lower than lowland due to water background. |
| Ripening / maturity | 115 – 145 | 0.35 – 0.60 | Longer season overall. |
Sudden NDVI drops on flood-prone fields often reflect a flood pulse, not crop damage. Cross-check with the **NDWI** layer and recent rainfall before flagging a stress event.
## Rainfed upland
Aerobic soil, no standing water, direct-seeded. Stands are sparser, canopy shorter, and NDVI rarely saturates. This is the most variable ecosystem to interpret from satellite.
| Phase | Days after sowing | Healthy NDVI | Notes |
| ---------------------------- | ----------------- | ------------ | ---------------------------------------------- |
| Emergence / early vegetative | 0 – 25 | 0.15 – 0.30 | Dry soil background, no water reflectance. |
| Tillering | 25 – 55 | 0.30 – 0.50 | Lower than lowland at the same phase. |
| Panicle initiation / booting | 55 – 80 | 0.45 – 0.65 | Peak biomass much lower than lowland. |
| Heading / flowering | 80 – 105 | 0.55 – 0.75 | Rarely saturates. NDRE gains are smaller here. |
| Ripening / maturity | 105 – 130 | 0.30 – 0.55 | Decline can be sharp if terminal drought hits. |
For direct-seeded fields (common in rainfed lowland and upland), shift each phase window by roughly 7 – 10 days later compared with transplanted paddy of the same variety.
## Main season and off season are not the same crop
Malaysian paddy runs two crops a year, and Malaysian agricultural statistics report them as two seasons — **main** and **off** — rather than as one annual figure. They differ in water regime, in radiation, and in how much of the water supply is under irrigation control: the off season falls in the Southwest Monsoon and leans harder on controlled irrigation than the main season does.
The seasons also sit on different calendars either side of the South China Sea. Peninsular paddy is planted **September to October** and harvested **December to March**. Sarawak is planted **October to November** and harvested **March to April** — roughly a month later at both ends.
**Compare main against main, and off against off.**
A main-against-off comparison measures the difference between two seasons, not the difference between two years of management. The gap between them is structural, so a field that reads lower in the off season has not necessarily got worse. The platform does not offer that comparison.
This does **not** rule out the chronic-zone work the diagnosis pages describe. A weak patch that repeats in the same place season after season is a real and useful signal, because it is being compared against the rest of the same field in the same season. What is refused is putting a main-season figure and an off-season figure on the same axis and calling the difference a change.
Season structure and the planting calendar follow the [FEWS NET Malaysia Country Book](https://help.fews.net/fde/v3/malaysia-country-book), which draws its crop statistics from *Perangkaan Agromakanan*, published by the Ministry of Agriculture and Agro-based Industry.
## Choosing the right ecosystem on a field
From the field overview, open **Settings** in the top-right.
Under **Crop**, choose **Paddy**. Additional paddy options appear.
Under **Ecosystem**, choose **Irrigated lowland**, **Rainfed lowland**, **Flood-prone / deepwater**, or **Rainfed upland**. The Health tab legend switches to the corresponding band table above.
Choose **Transplanted** or **Direct-seeded** and enter the date. The platform derives the current phase from days-after-establishment and applies the matching band automatically.
## Cross-ecosystem tips
* NDVI saturates above \~0.8 in every ecosystem. At heading and flowering, switch to **NDRE** regardless of ecosystem.
* Water background lowers NDVI early in the season for all flooded ecosystems. Do not read early-vegetative values against the generic scale.
* After one full season of history, prefer **Anomaly** view over the absolute bands. It adapts to your specific ecosystem, variety, and region.
* If a field straddles two ecosystems (for example, an upper terrace that is upland and a lower terrace that is rainfed lowland), split it into two fields so each gets the correct band.