NDVI - Vegetation Health

Land Usev04.00 credits/per_km2

Responsible:Niccolò Calandri

Layer NDVI - Vegetation Health on real sites

Real Estate UK

NDVI - Vegetation Health

Catania

NDVI - Vegetation Health

Berlino Quarter

NDVI - Vegetation Health

Mina el Cerrejon

NDVI - Vegetation Health

Overview

The Normalized Difference Vegetation Index (NDVI) is a spectral index that measures vegetation health and density derived from Sentinel-2 satellite imagery. Values range from -1 to +1, where higher values indicate healthier, denser vegetation.

NDVI exploits the difference in reflectance between near-infrared (strongly reflected by healthy vegetation) and red (absorbed by chlorophyll) wavelengths. It is one of the most widely used remote sensing indicators for assessing ecosystem health, land cover changes, and the presence of green areas over any user-selected period.

The platform shows NDVI in three complementary views: a trend line chart showing monthly variation, a circle gauge comparing site (ROI) with a control area (CA), and map layers showing spatial distribution at pixel level.

Calculation Methodology

NDVI is computed from ESA Sentinel-2 Level-2A surface reflectance imagery using the standard normalized difference of the near-infrared and red bands:

\text{NDVI} = \frac{NIR - RED}{NIR + RED}

where NIR = Band 8 (842 nm) and RED = Band 4 (665 nm), both at 10 m native resolution

Data are sourced from the ESA Sentinel-2 SR Harmonized collection at 10 m spatial resolution, available from April 2017 to present. Cloud-masked, water-masked pixels are excluded before computing the pixel-wise mean across all valid observations in the selected period. The analysis period is user-selectable and is not limited to annual frequency.

What You See on the Platform

Assessment > Land Use > "NDVI - Vegetation Health" (Line Chart)

Line Chart. A monthly time series chart displaying NDVI values for the site (ROI) and, optionally, the control area (CA).

Purpose: To show how vegetation health evolves across months, revealing seasonal patterns and year-to-year trends.

Description: The X-axis shows months (Jan--Nov or the full selected period). The Y-axis shows NDVI values from -1 to +1. Each monthly data point is drawn as a colored dot whose color reflects its NDVI value using the 21-color vegetation gradient (dark red = low/no vegetation; green = dense healthy vegetation). The solid line connects all site dots. A dashed grey line shows the control area (CA) values when the CA toggle is active. Below the chart, three statistics are shown: Min, Med (median), and Max.

How it's calculated: For each month, all cloud-free Sentinel-2 pixels within the site polygon are averaged. The resulting per-pixel mean is then spatially aggregated to produce the monthly ROI and CA values shown on the chart.

Legend: Dot color follows the continuous NDVI gradient:

Range	Color	Vegetation
-1.0 to 0.0	■ `#4b0000`	Bare soil, water, no vegetation
0.0 to 0.4	■ `#ff6347`	Very sparse to sparse vegetation
0.4 to 0.7	■ `#ffc100`	Moderate vegetation
0.7 to 1.0	■ `#008000`	Dense to very dense vegetation

Interpretation example:

If the chart shows a summer peak of 0.72 (green dot) and a winter low of 0.18 (orange dot), the site has strong seasonal vegetation dynamics typical of temperate grasslands or deciduous woodland -- high photosynthetic activity in summer, dormancy in winter.

Assessment > Land Use > "NDVI" (Gauge)

Gauge. A semicircular dual-arc gauge comparing the annual mean NDVI of the site (ROI) against the control area (CA).

Purpose: To summarise in a single visual comparison whether the site's vegetation health is better or worse than its surrounding control area.

Description: The gauge shows two concentric arcs: the outer arc represents the site (ROI) mean NDVI and the inner arc represents the control area (CA) mean NDVI. Both arcs are coloured according to the A--E quality scale. The center displays the ROI value and the CA value, plus a delta badge showing the difference (ROI - CA). A positive delta (green badge) means the site has healthier vegetation than its surroundings; negative (red badge) means the opposite.


0.62NDVI
-11

How it's calculated: The annual mean NDVI is computed from all monthly pixel-wise means within the selected period. The ROI arc color corresponds to the quality grade assigned based on the thresholds.

Legend:

Level	NDVI Range	Color	Meaning
A	0.8 -- 1.0	■ `#00A67A`	Very dense, healthy vegetation
B	0.5 -- 0.8	■ `#00DF80`	Dense vegetation, healthy forests
C	0.2 -- 0.5	■ `#FFD21E`	Grasslands, crops, moderate vegetation
D	0.1 -- 0.2	■ `#FF8B16`	Sparse vegetation, degraded areas
E	-1.0 -- 0.1	■ `#FF367F`	Water, bare soil, artificial surfaces

Interpretation example:

If the gauge shows ROI = 0.62 (grade B) and CA = 0.45 (grade C), the site has denser and healthier vegetation than its surrounding landscape -- a positive ecological signal indicating the site supports greater plant biomass than its regional context.

Assessment > Land Use > "NDVI" (Assessment Sidebar Row)

Assessment Sidebar Row. A single row in the Land Use KPI panel within the Assessment sidebar, showing the current NDVI grade.

Purpose: To give a quick A--E quality rating for NDVI as part of the broader land use assessment.

Description: The row shows the KPI category label "NDVI", the latest numeric value, and a colored grade badge (A--E). Clicking the row navigates to the full NDVI section.

How it's calculated: The grade is assigned using the same thresholds as the gauge above (A >= 0.8, B 0.5--0.8, C 0.2--0.5, D 0.1--0.2, E < 0.1).

Interpretation example:

If the sidebar shows an orange "D" badge next to NDVI 0.15, the site has sparse vegetation -- consistent with degraded land, heavily impervious surfaces, or low-density scrub.

Maps > Land Use > "NDVI" (Map Layer)

Map Layer. An interactive satellite map layer displaying NDVI values at pixel level across the site and surroundings.

Purpose: To show the spatial distribution of vegetation health, identifying which parts of the site have the most (or least) productive vegetation.

Description: Each pixel is coloured according to its NDVI value using the continuous 21-color gradient. Dark red pixels indicate bare soil, water, or very low vegetation. Green pixels indicate dense, healthy vegetation. The site boundary polygon is overlaid. Users can switch between three sub-layers: standard NDVI (single period), NDVI annual range (seasonal range for the year), and NDVI range delta (change between two periods).

NDVI Scale
-1
+1
Bare soil / waterDense healthy vegetation

How it's calculated: The map tile is generated from the Sentinel-2 pixel-wise mean composite for the selected period. Each pixel's NDVI value is mapped to a color from the 21-step gradient (vmin = -1, vmax = +1).

Legend: Continuous gradient from ■ #4b0000 (NDVI = -1) to ■ #008000 (NDVI = +1):

Range	Color	Vegetation
-1.0 to 0.0	■ `#4b0000` to ■ `#8b0000`	Bare soil, water, no vegetation
0.0 to 0.4	■ `#8b0000` to ■ `#ff6347`	Very sparse to sparse vegetation
0.4 to 0.7	■ `#ff6347` to ■ `#ffc100`	Moderate vegetation
0.7 to 1.0	■ `#d4f517` to ■ `#008000`	Dense to very dense vegetation

Interpretation example:

If the map shows a predominantly green interior with dark red patches along roads and paved areas, the site has a healthy natural core with impervious-surface fragmentation -- a common pattern in urban nature reserves.

Maps > Land Use > "NDVI annual range" (Map Layer)

Map Layer. A map layer showing the average NDVI over a full annual range, typically from the previous autumn through the following summer (configurable in site settings).

Purpose: To represent the site's annual vegetation cycle rather than a single snapshot, reducing the effect of short-term anomalies.

Description: Same color gradient as the standard NDVI layer. The range period is defined in the site's layer generation settings (default: October of the previous year to June of the following year). The map tooltip shows: "Average NDVI in the period {start} -- {end}".

How it's calculated: All cloud-free Sentinel-2 images within the configured range period are composited into a pixel-wise mean.

Legend: Same 21-color continuous gradient as the standard NDVI layer (■ #4b0000 to ■ #008000).

Interpretation example:

An annual range layer showing NDVI 0.65 across most of the site confirms consistently healthy vegetation across seasons -- a more robust indicator than a single summer reading.

Maps > Land Use > "NDVI range delta between years" (Map Layer)

Map Layer. A map layer showing the pixel-wise difference in NDVI between two annual range composites, indicating vegetation change over time.

Purpose: To identify where vegetation has improved, degraded, or remained stable between two comparable periods.

Description: Pixels are coloured using an 11-color diverging gradient: red tones indicate vegetation loss, yellow/neutral indicates no change, green tones indicate vegetation gain. The tooltip shows: "Difference between average NDVI in the periods {x1}--{y1} and {x2}--{y2}".

How it's calculated: The delta is computed pixel-wise as (NDVI of final period) - (NDVI of initial period). The result ranges from approximately -0.25 to +0.25.

Legend: Diverging gradient from ■ #8B0000 (strong vegetation loss) through ■ #FFFF00 (no change) to ■ #008000 (strong vegetation gain).

Interpretation example:

If a large green area appears on the delta layer where last year showed bare soil, the site has undergone significant reforestation or grassland recovery -- a positive change worth monitoring in subsequent years.

Data Sources

Source	Provider	Coverage	Resolution	Period
ESA Sentinel-2 SR Harmonized	ESA / Copernicus	Global	10 m (Band 4, Band 8)	2017-04-01 -- present

Detailed Methodology

All Sentinel-2 L2A images within the requested date range and site area are acquired. Cloud masking is applied using the Scene Classification Layer (SCL): pixels corresponding to cloud shadow, medium/high cloud probability, and thin cirrus are excluded. Water pixels are excluded by default.

NDVI is calculated per pixel as the normalized difference of Band 8 (near-infrared, 842 nm) and Band 4 (red, 665 nm), both at 10 m native resolution. Temporal compositing produces a pixel-wise mean across all valid observations in the period. Spatial aggregation then computes min, max, and mean statistics separately over the site (ROI) and the control area (CA).

For the map layers, the pixel-wise mean composite is rendered using a 21-step color gradient (vmin = -1, vmax = +1). The annual range layer composites all cloud-free images within the configured range period (default: October -- June). The delta layer computes pixel-wise differences between two annual range composites.

Map resolution defaults to 15 m; automatically upgraded to 20 m for sites larger than 100 km² and 25 m for sites larger than 200 km².

Technical Notes

NDVI map layers must be generated from satellite imagery before the KPI values can be computed; generation is triggered manually or on schedule from site settings
NDVI tends to saturate for very dense vegetation (values above 0.8--0.9); the Enhanced Vegetation Index (EVI) is a better alternative in tropical forest sites
The index is sensitive to residual atmospheric effects, solar angle, and soil brightness -- these factors can introduce variability unrelated to actual vegetation changes
Water masking is enabled by default, so water bodies are excluded from the spatial statistics; pixels over rivers, lakes, or flooded areas appear as missing data
Data availability begins from April 2017 (Sentinel-2 L2A archive start date); sites requiring pre-2017 analysis cannot be served by this KPI
Cloud cover can reduce the number of valid observations per month, especially in winter or tropical wet seasons; months with very few clear observations may show less reliable values
Related KPIs: Land Use (habitat classification), LST (Land Surface Temperature), and Flora (AI-derived plant diversity) all use NDVI as an input or complement

Bibliography

Rouse, J.W., Haas, R.H., Schell, J.A., Deering, D.W. (1974). "Monitoring vegetation systems in the Great Plains with ERTS". NASA Special Publication, 351, 309--317. NASA Technical Reports
Tucker, C.J. (1979). "Red and photographic infrared linear combinations for monitoring vegetation". Remote Sensing of Environment, 8(2), 127--150. DOI: 10.1016/0034-4257(79)90013-0
ESA Copernicus. (2021). Sentinel-2 User Handbook. European Space Agency. ESA Sentinel-2