Managing soil via satellites and GIS

The South East Rivers Trust is using satellite image processing and Geographical Information Systems (GIS) to target measures to improve soil management and water quality, writes Dr Alastair Pearson, one of our GIS analysts.

Methods of soil management include the introduction of cover crops to promote better nutrient balance and soil structure, improve weed control and biodiversity, and the reduction of erosion.

Charting seasonal variations

Traditional land cover maps are excellent data sources for identifying different types of cover such as arable, improved grassland, broadleaved and coniferous woodland etc. However, they offer little on the condition of these land cover types.

SERT is interested in knowing the seasonal variation and, in particular, the duration of exposure of the soil to the effects of wind and rain within a river catchment.

The effects of wind and water on exposed soil during the winter months are well documented: they lead to the loss of organic matter, nitrates, phosphorous and pesticides into water courses with subsequent deterioration in water quality and biodiversity.

SERT uses Sentinel-2 satellite imagery from the European Space Agency (ESA) at 10 metre resolution (each pixel or reflectance value represents an area of 10x10m).

This provides a measure of green vegetation, applying a simple mathematical combination of the visible red and the near-infrared (NIR) bands to detect the presence and condition of green vegetation using the normalised difference vegetation index (NDVI):

NDVI calculation for GIS

A satellite image of the south east of England. Contains modified Copernicus Sentinel data 2021 processed by Sentinel Hub

Building a picture of a year's data

NDVI analysis essentially creates a new image based on the vegetation cover characteristics and reflectance from the original image. As the imagery from Sentinel has a 10 metre spatial resolution, variations within fields, for example the presence of woody hedgerows, are reflected in the output.

As vegetation cover varies throughout the year, several images spread through a 12-month period are used to calculate a mean NDVI value.

The final layer therefore reflects seasonal variations in all habitat types as well as performing an important habitat condition indicator.

The six images in the graphic show the variation in green vegetation cover over a 12-month period in an area of the Medway catchment. The areas in red indicate very low green vegetation cover or bare soil whereas the areas in dark green indicate relatively healthy vegetation cover.

If we simply calculate the mean NDVI values, then we can obtain a fairly accurate indication of the vegetation cover over a 12-month period and of course identify areas that are at greater risk of exposure to the effects of sun, wind and rain.

Six NDVI images over a 12-month period. Contains OS data © Crown Copyright and database right 2020

Helping farmers manage land

The NDVI mapping forms an important focal point in discussing land management practices with farmers as well as providing invaluable ground truthing of the maps themselves.

Other layers of data such as slope gradient, land cover type, soil texture and infiltration rates relevant to our study of potential risks to water quality and water supply are included in our final analysis.

The approach outlined here currently forms an integral part of our analysis of water resources in the Beult catchment for Southern Water and in the Lea and Colne catchments for Affinity Water.

For more information about our GIS work, email

Mean NDVI values over a 12-month period. Contains OS data © Crown Copyright and database right 2020