Skip to content

Sentinel-2 NDVI

What this dataset is

The Sentinel-2 Level-2A constellation provides multispectral surface reflectance at 10–60 m resolution with a ~5-day revisit time over global land masses. CubeDynamics derives the Normalized Difference Vegetation Index (NDVI) from the red (B04) and near-infrared (B08) bands at 10 m resolution, producing a (time, y, x) vegetation cube over the requested window.

Quickstart

Get the stream (CubeDynamics grammar)

import cubedynamics as cd
from cubedynamics import pipe, verbs as v

cube = cd.ndvi(
    lat=40.0,
    lon=-105.25,
    start="2023-06-01",
    end="2023-07-01",
)

pipe(cube) | v.mean(dim="time") | v.plot()

Preview plot

Sentinel-2 NDVI preview

!!! note Image placeholder — after running the code below locally, save a screenshot to docs/assets/datasets/sentinel2_ndvi-preview.png.

Regenerate this plot

  1. Execute the Quickstart snippet to stream a month of NDVI over the example location.

  2. Capture the CubePlot from the pipeline:

    python viewer = (pipe(cube) | v.mean(dim="time") | v.plot()).unwrap() viewer.save("docs/assets/datasets/sentinel2_ndvi-preview.html")

  3. Open docs/assets/datasets/sentinel2_ndvi-preview.html in a browser and save a 1200×700 px PNG screenshot to docs/assets/datasets/sentinel2_ndvi-preview.png.

Who collects it and why

The European Space Agency (ESA) and the European Commission operate Sentinel-2 to deliver routine optical imagery for land monitoring, vegetation status, and disaster response. The atmospherically corrected Level-2A product is widely used for vegetation phenology and ecosystem monitoring, making it an authoritative source for NDVI analyses.

How CubeDynamics accesses it

Sentinel-2 scenes are streamed remotely through the cubo API, which signs and reads cloud-optimized GeoTIFF assets without downloading entire archives. NDVI is computed on-the-fly from the requested bands, and the resulting cube remains lazily evaluated so downstream verbs trigger IO only as needed. Long requests can be chunked temporally to avoid large STAC queries while preserving the VirtualCube-style streaming behavior.

Important variables and dimensions

Field Meaning Units
time Observation timestamp ISO date
y / x Spatial coordinates in the native UTM projection meters
NDVI (NIR − Red) / (NIR + Red) vegetation index unitless
band (optional, when returning raw bands) Reflectance bands such as B04 (red) and B08 (NIR) unitless reflectance

Citation

Didan, K. (2015). MOD13Q1 MODIS/Terra Vegetation Indices (Version 6). NASA EOSDIS Land Processes DAAC. https://doi.org/10.5067/MODIS/MOD13Q1.006

Back to Datasets Overview
Next recommended page: Which dataset should I use?