Index Library

Geospatial Index Library

Comprehensive reference for spectral, environmental, and terrain indices. Every index includes its mathematical formula, physical interpretation, practical applications, and known limitations.

ndvi_calculation.py

python

import numpy as np
import rasterio

# Open the Red and NIR bands
with rasterio.open('sentinel2_B04.tif') as red_src:
    red = red_src.read(1).astype('float32')

with rasterio.open('sentinel2_B08.tif') as nir_src:
    nir = nir_src.read(1).astype('float32')

# Avoid division by zero
np.seterr(divide='ignore', invalid='ignore')

# Compute NDVI
ndvi = (nir - red) / (nir + red)

# Handle NaNs
ndvi = np.nan_to_num(ndvi, nan=-1.0)

Vegetation Indices

NDVI

Normalized Difference Vegetation Index

Interpretation

Values range from −1 to +1. Dense healthy vegetation yields values 0.6–0.9. Bare soil ~0.1–0.2. Water typically negative.

Applications

Crop health monitoring, deforestation detection, phenology tracking, drought assessment.

Limitations

Saturates in dense canopies (LAI > 3). Sensitive to soil background in sparse vegetation. Atmospheric effects require correction.

ndvi.py

python

ndvi = (nir - red) / (nir + red)

EVI

Enhanced Vegetation Index

Interpretation

Optimised for high-biomass regions. Standard coefficients: G=2.5, C₁=6, C₂=7.5, L=1. Less prone to saturation than NDVI.

Applications

Tropical forest monitoring, high-biomass agricultural areas, canopy structural analysis.

Limitations

Requires blue band (not available on all sensors). More complex computation.

evi.py

python

evi = 2.5 * ((nir - red) / (nir + 6 * red - 7.5 * blue + 1))

SAVI

Soil Adjusted Vegetation Index

Interpretation

L is a soil brightness correction factor (typically 0.5). Reduces soil background noise for sparse vegetation.

Applications

Arid/semi-arid regions, early crop growth stages, rangeland monitoring.

Limitations

Requires empirical selection of L. Less effective for dense vegetation.

savi.py

python

L = 0.5
savi = ((nir - red) / (nir + red + L)) * (1 + L)

GNDVI

Green NDVI

Interpretation

More sensitive to chlorophyll concentration than NDVI. Useful for assessing nitrogen status.

Applications

Precision agriculture, chlorophyll estimation, crop nitrogen management.

Limitations

Less documented in literature compared to NDVI. Can saturate at very high chlorophyll levels.

gndvi.py

python

gndvi = (nir - green) / (nir + green)

VIIRS

Visible Infrared Imaging Radiometer Suite

Interpretation

Provides radiometric data from the VIIRS instrument aboard Suomi-NPP and NOAA-20.

Applications

Nighttime lights, fire detection, ocean color, sea surface temperature.

Limitations

Coarser resolution than Sentinel-2 or Landsat for land applications.

viirs.py

python

# Pseudo-code for fetching VIIRS data
viirs_data = client.get_viirs(band='DNB', date='2026-05-01')

Water Indices

NDWI

Normalized Difference Water Index

Interpretation

Positive values indicate water bodies. Originally proposed by McFeeters (1996).

Applications

Surface water mapping, flood extent delineation, wetland monitoring.

Limitations

Can be confused by built-up areas. Modified NDWI (MNDWI using SWIR) often preferred.

ndwi.py

python

ndwi = (green - nir) / (green + nir)

MNDWI

Modified NDWI

Interpretation

Better discrimination between water and built-up areas than NDWI. Positive for water, negative for built-up.

Applications

Urban water body mapping, improved flood mapping in built environments.

Limitations

Requires SWIR band. Mixed pixels at water-land boundaries.

mndwi.py

python

mndwi = (green - swir) / (green + swir)

Soil & Fire Indices

NBSI

Normalized Bare Soil Index

Interpretation

Highlights bare soil by exploiting reflectance contrast between SWIR and visible green bands.

Applications

Desertification monitoring, mining reclamation, construction site detection.

Limitations

May confuse dry vegetation with bare soil in arid environments.

nbsi.py

python

nbsi = (swir1 - green) / (swir1 + green)

NBR2

Normalized Burn Ratio 2

Interpretation

Sensitive to changes in vegetation moisture and burn severity using two SWIR bands.

Applications

Burn severity assessment, post-fire recovery monitoring, wildfire damage mapping.

Limitations

Less effective under cloud cover. Requires both SWIR bands.

nbr2.py

python

nbr2 = (swir1 - swir2) / (swir1 + swir2)

NDI7

Normalized Difference Index 7

Interpretation

Sensitive to moisture content in both vegetation and soil. Useful for drought monitoring.

Applications

Fuel moisture estimation, fire risk assessment, drought mapping.

Limitations

Atmospheric correction critical for SWIR bands. Limited to sensors with SWIR2 band.

ndi7.py

python

ndi7 = (nir - swir2) / (nir + swir2)

Terrain Indices

TWI

Topographic Wetness Index

Interpretation

Where a = upslope contributing area per unit contour length and β = local slope angle. Higher TWI indicates areas prone to saturation.

Applications

Hydrological modelling, soil moisture prediction, wetland delineation.

Limitations

Assumes steady-state conditions. Sensitive to DEM resolution and flow direction algorithm.

twi.py

python

import numpy as np
twi = np.log(upslope_area / np.tan(slope))

TPI

Topographic Position Index

Interpretation

Difference between elevation at a point and mean elevation of surrounding neighbourhood. Positive = ridge, negative = valley.

Applications

Landform classification, habitat mapping, erosion risk assessment.

Limitations

Results depend heavily on neighbourhood size selection.

tpi.py

python

import scipy.ndimage as nd
mean_z = nd.uniform_filter(elevation, size=3)
tpi = elevation - mean_z

Working towards it: Additional indices (ARVI, MSAVI2, BSI, NDBI, UI, NDSI, TSAVI, WDRVI, CIgreen, CIred-edge) and full terrain analysis suite (curvature, roughness, hillshade, flow accumulation) are being documented.