NDVI Calculator Online: Sentinel-2 NDVI Map for Fields
Draw your field or parcel on the map. NDVI is computed from Sentinel-2 satellite data, cloudy pixels are removed, and you get class areas and a chart of recent change.
Summary
Click a coloured pixel on the map to see its value and SCL class. Grey pixels were excluded because of cloud, cloud shadow or snow.
Area by class
Areas are the number of valid pixels multiplied by the pixel area (10 m x 10 m = 100 m²). 1 hectare = 10,000 m² = about 2.47 acres.
Time series: area mean
Show as table
The time series is computed at 20 m resolution to reduce data transfer. Only scenes in which at least 80% of the area is cloud-free are used.
Data Source and Privacy
Data: Copernicus Sentinel-2 L2A via AWS Open Data. To search for scenes, the corner coordinates of the area you draw, the date range and the cloud threshold are sent to the Element 84 Earth Search service. Only the part of each band file that covers your area is downloaded from AWS, and the calculation runs in your browser. Place search text goes to OpenStreetMap Nominatim.
Related Tools
What NDVI is and what it is used for
NDVI (Normalized Difference Vegetation Index) relies on the fact that plants absorb red light and strongly reflect near-infrared light. The formula is NDVI = (NIR - Red) / (NIR + Red) and the result ranges from -1 to +1. Green, leafy vegetation gives high values; bare soil, rock and buildings give low values, and water is usually negative.
With this NDVI calculator you do not need to download satellite images or install software. When you draw the boundary of a field, the tool finds the most recent suitable scene from the Sentinel-2 satellites of the European Union's Copernicus programme, removes cloudy pixels and draws the NDVI map. It can be used to monitor crop health, spot weaker zones inside a parcel, follow sowing and harvest, or track change in pasture and forest.
How to use it
- Find your field: zoom the map or type a village, town or "latitude, longitude" in the search box.
- Draw the field boundary as a rectangle or polygon with the tools at the top left or the buttons in the panel. The bounding box of the area can be up to 25 km².
- The tool searches Sentinel-2 scenes in the selected date range and calculates the most recent scene that is cloud-free over your area. You can pick another date from the list, or change the date range and scene cloud threshold and search again.
- The summary shows the mean, median, minimum and maximum; the table lists the area of each class in hectares. Click a pixel to see its value.
- The "Chart" button plots the mean NDVI of the area over the most recent clear scenes as a time series.
- Download the map as PNG, the NDVI values as GeoTIFF and the time series as CSV.
How to read NDVI values
- Below 0: water, cloud shadow, snow or wet dark surfaces.
- 0 to 0.2: bare soil, fallow, harvested fields, rock and built-up land.
- 0.2 to 0.4: sparse vegetation, emerging crops, drying pasture or stressed plants.
- 0.4 to 0.6: moderately dense, developing vegetation.
- Above 0.6: dense, vigorous vegetation such as a fully developed crop or forest.
These ranges are a general guide. Crops such as wheat, maize or sunflower give very different values depending on growth stage. Rather than looking at a single date, comparing a field with its own time series and with neighbouring fields of the same crop gives a more reliable picture.
Two more indices are available. NDWI (B03 and B08) separates open water and wet surfaces; values above 0 are usually water. NDMI (B08 and B11) reflects leaf water content; low values can point to water stress.
Data and method
Scenes are searched in the sentinel-2-l2a collection of the Element 84 Earth Search STAC service. The L2A product is atmospherically corrected surface reflectance. Bands are read from Cloud Optimized GeoTIFF files in the AWS open data registry; only the window that covers your area is downloaded, not the whole file. The red (B04) and near-infrared (B08) bands have 10 m resolution. Values are converted to reflectance with the scale factor of the scene; the +1000 offset in scenes processed since 2022 is already removed in the Earth Search data, so it is not applied twice.
Clouds are masked with the 20 m SCL (scene classification) band: cloud, cloud shadow, thin cirrus, snow, saturated and empty pixels are excluded. The time series is computed at 20 m to reduce data transfer and uses scenes in which at least 80% of the area is cloud-free.
Libraries used: geotiff.js 2.1.3 (MIT), Leaflet 1.9.4 (BSD-2-Clause), Leaflet.draw 1.0.4 (MIT), proj4js 2.9.0 (MIT), Chart.js 4.4.1 (MIT) and Lucide icons (ISC). The GeoTIFF export is written by a small built-in writer. Sentinel-2 data can be used freely under the Copernicus data policy; exports carry the attribution "Contains modified Copernicus Sentinel data".
Accuracy limits
- Resolution: one pixel is 10 m by 10 m. In narrow fields and along field edges, roads, tree rows or the neighbouring parcel mix into the result.
- Cloud and haze: the SCL mask does not always catch thin cloud, haze and cloud edges. If values look unexpectedly low, check another date.
- Season and crop: values depend on crop type, sowing date, irrigation and sun angle. There is no fixed "healthy" threshold.
- Field visit: NDVI shows where a problem may be, not why. Before deciding on fertiliser, spraying or irrigation, inspect the field and, if needed, have soil or leaf samples analysed.
For organisations
Regular monitoring of many parcels, crop mapping and yield estimation need a managed workflow rather than one-off checks. If your organisation needs satellite and drone image analysis inside an ArcGIS environment, learn more about Rodosto Teknoloji or get in touch with us.
Frequently Asked Questions
What NDVI value means healthy vegetation?
There is no single threshold. Roughly, values above 0.6 indicate dense, vigorous vegetation, 0.2 to 0.4 sparse vegetation and values below 0.2 bare soil. The same crop gives different values at different growth stages, so comparing a field with its own time series and with neighbouring fields is more meaningful.
How often are new Sentinel-2 images available?
At mid latitudes the Sentinel-2 satellites typically image the same place every 2 to 5 days. Cloudy days are masked out, so the number of usable scenes depends on the season and the region.
Is the result reliable for a small garden or a field of one or two acres?
A Sentinel-2 pixel covers 10 m by 10 m, or 100 m². A field of one or two acres covers only a few dozen pixels, and edge pixels are influenced by neighbouring land. The result shows the general pattern; decisions at plant level need a field visit.
Which programs can open the GeoTIFF export?
The export is a single-band 32-bit floating point (float32) GeoTIFF in the UTM coordinate system of the scene (for example EPSG:32636). Pixels outside the area or under cloud are set to -9999. It opens in common GIS software, including ArcGIS Pro.