# Use Case: Variable-Rate Nitrogen (VRA) for Potatoes to Realize 5–10% More Yield

**Read the full story:** [Realize 5–10% more yield in potatoes (case study)](https://geopard.tech/blog/vk2kgcd5o1-realize-5-10-more-yield-in-potatoes/)

<figure><img src="https://3272281156-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYICBELdyAXXebKAzfLOR%2Fuploads%2FArzijS5zB8W3t8MlSq4U%2Fimage.png?alt=media&#x26;token=e1ab6d2d-8140-4349-8cf5-d9f75534159c" alt="" width="375"><figcaption></figcaption></figure>

## Summary

**Client:** [**Databoerin (NL)** ](https://databoerin.nl/)— an innovative smart-farming consultancy led by **Nicole Bartelds**. They specialize in potato fertilization plans and have used GeoPard for several years to create **variable-rate side-dress nitrogen** based on **Sentinel-2 RedEdgeChlorophyll** and a crop growth model. Reported outcomes from multi-year fieldwork: **+5–10% yield** with the **same total N**, more even canopy, and cleaner, more uniform ripening. Typical program: \~**60% base N** early, then **side-dress at crop closure** guided by chlorophyll-derived N uptake.

Optimize **side-dressing nitrogen** in potatoes using **Sentinel-2 chlorophyll (RedEdgeChlorophyll)**, a crop growth model, and GeoPard VRA prescriptions (ISOXML). Field results reported by a Dutch consultancy show **+5–10% yield** with the **same total N**, redistributed by need.

<figure><img src="https://3272281156-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYICBELdyAXXebKAzfLOR%2Fuploads%2FCf4ogCbPy8oSWRMMoNPH%2Fimage.png?alt=media&#x26;token=c7ec8b21-34f3-42bb-9a9e-a33fb6a29f37" alt="Side-dressing VRA N map for the fertilizer spreader"><figcaption></figcaption></figure>

### When to Use

* Potatoes on soils with **variable mineralization** or uneven vigor.&#x20;
* Growers/consultants/dealers aiming to **redistribute N** at side-dress timing (row closure).&#x20;

### Data Inputs

* **Satellite imagery:** Sentinel-2; GeoPard processes **cloud-free images within \~1 day**, exposing **RedEdgeChlorophyll**.&#x20;
* **Field boundaries & zones:** from farm/FMS or created in GeoPard.
* Optional: soil tests, prior yield, topography for context.

### Agronomic Logic (Potatoes, N side-dressing)

* **Base application:** early season \~**60%** of the recommended dose (manure + mineral N).&#x20;
* **Side-dressing at crop closure:** use chlorophyll-based N uptake to **add N only where needed**.&#x20;

***

### Step-by-Step (GeoPard Workflow)

1. **Ingest imagery**\
   Use the latest **cloud-free Sentinel-2** scene. GeoPard provides **RedEdgeChlorophyll** for each field.&#x20;
2. **Estimate N uptake**\
   From **chlorophyll content** (proxy for N in leaves) estimate current N uptake vs. **optimal uptake** from the crop model.
3. **Compute the delta**\
   Convert the gap to **variable rates (kg N/ha)** per management unit.
4. **Create the VRA map, Use** [**Rates Distribution Tool**](https://docs.geopard.tech/geopard-tutorials/product-tour-web-app/zones-maps-and-analytics/assign-variable-rates-in-the-zones-ag-inputs-rates-distribution-tool)\
   Set **min/max rates** and operational constraints, then **export ISOXML** for the spreader/terminal.&#x20;
5. **Apply & verify**\
   Execute side-dressing; monitor canopy evenness and later **yield**. Use **as-applied** + new imagery to validate.

***

### Outputs

* **RedEdgeChlorophyll map** (canopy N proxy).
* **VRA N side-dressing prescription** (ISOXML).
* Optional **EVI2** to visualize variety-driven differences and canopy evenness.

### Expected Impact

* **Yield:** reported **+5–10%** with unchanged total N, thanks to better spatial allocation.&#x20;
* **Crop uniformity:** more even canopy → more uniform **phytophthora** protection and **ripening**.&#x20;

***

### Best Practices

* Time side-dressing **around crop closure**.
* Exclude headlands/borders; set **rate floors/ceilings** for machine safety.
* Watch [**clouds & shadows**](https://geopard.tech/blog/cloud-and-shadow-detection-in-agriculture/); confirm units (kg/ha).
* If manure mineralization is high, keep conservative max rates and validate with **test strips**.

### Validation & KPIs

* **Canopy evenness** (post-application imagery)
* **Yield uplift** by zone vs. baseline.
* [**Nitrogen Use Efficiency (NUE)**](https://docs.geopard.tech/geopard-tutorials/agronomy/nitrogen-use-efficiency-nue-and-nitrogen-uptake) and % area within target rate band.

***

### Troubleshooting

* **No recent imagery?** Wait for next cloud-free pass; re-run the map.
* **Patchy vigor due to variety?** Use **EVI2** to separate variety effects from N signal.
* **Controller rejects file?** Re-export ISOXML; check terminal profile & coordinate system.

***

### FAQ

**Which index is used?**\
**RedEdgeChlorophyll** for N status; **EVI2** for vigor/variety contrasts.

**How much base N vs. side-dress?**\
Typical pattern in this workflow: \~**60% base**, remainder at side-dress informed by chlorophyll.&#x20;

**What uplift to expect?**\
Field experience cited: **+5–10% yield** with the same total N, redistributed. Results vary.

***

#### Attribution

Use-case adapted from a project with **Databoerin (NL)** using GeoPard analytics for potato fertilization.