Water & Soil Indices

MNDWI

Modified Normalized Difference Water Index. An enhancement of NDWI that uses the shortwave infrared (SWIR) band instead of NIR, providing better discrimination between water bodies and built-up areas in urban environments.

Formula

MNDWI = (Green - SWIR) / (Green + SWIR)

Overview

The Modified Normalized Difference Water Index (MNDWI) extracts open water features from satellite imagery while suppressing noise from built-up land. Developed by Hanqiu Xu in 2006 as an improvement over McFeeters' NDWI, it substitutes SWIR for NIR, exploiting water's stronger SWIR absorption and built-up surfaces' higher SWIR reflectance.

How It Works

MNDWI = (Green − SWIR) / (Green + SWIR). On Sentinel-2: (B3 − B11) / (B3 + B11). Water produces strongly positive values because it absorbs SWIR. Built-up areas produce negative values due to high SWIR reflectance. A threshold around 0.0–0.3 separates water from non-water.

Key Facts

  • Formula: (Green − SWIR) / (Green + SWIR)
  • Key advantage over NDWI: suppresses built-up area false positives.
  • Sentinel-2 bands: B3 (Green) and B11 (SWIR-1, 20 m resolution).

Applications

Urban Water Body Mapping

Excels at detecting water bodies within cities where NDWI misclassifies buildings as water.

Flood Extent Monitoring

Delineating inundated areas in mixed urban-rural floodplains.

Coastal Monitoring

Tracking shoreline changes and tidal flat exposure.

Limitations & Considerations

Can struggle with turbid water. Shadows from tall buildings may mimic water. The SWIR band on Sentinel-2 (20 m) limits spatial detail compared to the green band (10 m).

History & Background

Published by Hanqiu Xu in 2006 in the International Journal of Remote Sensing as a direct improvement to McFeeters' 1996 NDWI. Now the standard choice for urban water mapping.

Analyze MNDWI data with LYRASENSE

Use our agentic notebook environment to work with satellite data and compute indices like MNDWI — no setup required.