Water and Moisture Indices in GeoIndexR

Overview

Delineating water bodies and tracking moisture conditions are foundational tasks in hydrology, wetland monitoring, and irrigation management. GeoIndexR implements three key indices:

1. NDWI (Normalized Difference Water Index, McFeeters 1996)
2. MNDWI (Modified Normalized Difference Water Index, Xu 2006)
3. NDMI (Normalized Difference Moisture Index, Gao 1996)

1. NDWI (McFeeters, 1996)

McFeeters designed NDWI to delineate open water features by maximizing green band reflectance and minimizing NIR reflectance:

\[\text{NDWI} = \frac{\text{GREEN} - \text{NIR}}{\text{GREEN} + \text{NIR}}\]

Water bodies typically have positive NDWI values (\(\text{NDWI} > 0\)), while terrestrial vegetation and dry soil display negative values.

library(GeoIndexR)
img <- get_example_data()

ndwi <- geo_index(img, "NDWI")
index_summary(ndwi)
#> 
#> === GeoIndexR Spectral Summary ===
#> 
#>  index     min    max   mean  median    sd     q05     q25    q75   q95 na_pct
#>   NDWI -0.8455 0.7243 -0.228 -0.2392 0.513 -0.8015 -0.7592 0.3265 0.569      1
#>  total_cells
#>          100

2. MNDWI (Xu, 2006)

In urbanized and complex landscapes, built-up surfaces often produce false positive signals under classical NDWI. Xu (2006) replaced the NIR band with the Shortwave Infrared (SWIR) band:

\[\text{MNDWI} = \frac{\text{GREEN} - \text{SWIR}}{\text{GREEN} + \text{SWIR}}\]

Because built-up areas reflect strongly in SWIR, their MNDWI values are negative, clearly separating urban structures from open water.

mndwi <- geo_index(img, "MNDWI")
index_summary(mndwi)
#> 
#> === GeoIndexR Spectral Summary ===
#> 
#>  index     min    max    mean  median     sd    q05     q25   q75    q95 na_pct
#>  MNDWI -0.5776 0.8329 -0.0413 -0.3182 0.5172 -0.494 -0.4105 0.645 0.7964      1
#>  total_cells
#>          100

3. NDMI (Gao, 1996)

The Normalized Difference Moisture Index monitors vegetation liquid water content:

\[\text{NDMI} = \frac{\text{NIR} - \text{SWIR}}{\text{NIR} + \text{SWIR}}\]

Higher values indicate well-hydrated vegetation canopies, whereas low or negative values signal drought or water stress.

ndmi <- geo_index(img, "NDMI")
index_summary(ndmi)
#> 
#> === GeoIndexR Spectral Summary ===
#> 
#>  index     min    max   mean median     sd     q05     q25    q75    q95 na_pct
#>   NDMI -0.3523 0.6916 0.2813 0.4287 0.3275 -0.2766 -0.0806 0.5354 0.6531      1
#>  total_cells
#>          100

Multi-Index Stack & Comparison

We can compute and compare water and moisture indices together:

water_stack <- geo_indices(img, c("NDWI", "MNDWI", "NDMI"))
index_summary(water_stack)
#> 
#> === GeoIndexR Spectral Summary ===
#> 
#>  index     min    max    mean  median     sd     q05     q25    q75    q95
#>   NDWI -0.8455 0.7243 -0.2280 -0.2392 0.5130 -0.8015 -0.7592 0.3265 0.5690
#>  MNDWI -0.5776 0.8329 -0.0413 -0.3182 0.5172 -0.4940 -0.4105 0.6450 0.7964
#>   NDMI -0.3523 0.6916  0.2813  0.4287 0.3275 -0.2766 -0.0806 0.5354 0.6531
#>  na_pct total_cells
#>       1         100
#>       1         100
#>       1         100