Talk on Satellite-Based Crop Coefficients for Smarter Potato Irrigation by Prof M. Steyn

 

Precision Agriculture and Irrigation

How real-time satellite observations could help South African growers conserve water, reduce energy costs and protect potato yields

About the Speaker

Professor Martin Steyn of the University of Pretoria presented research on using satellite-derived crop coefficients to improve irrigation scheduling for potatoes under South African production conditions.

Professor Steyn was also the External Examiner for the MSc Agronomy programme at the Faculty of Agriculture, University of Mauritius. His visit provided an opportunity to share research insights with Faculty staff and strengthen academic engagement in precision agriculture and sustainable crop production.

Prof. Martin Steyn with the Programme Coordinator, Mr. M. Choonea and the Dean of the Faculty, Assoc prof J. Govinden Soulange.

Why Smarter Potato Irrigation Matters

Potatoes are an important South African crop, with approximately 50,000 hectares planted annually across 16 production regions. Around 85% of this area is irrigated, mainly using centre-pivot systems.

However, potatoes are temperate plants being produced in a generally water-scarce, semi-arid country. They are highly sensitive to water stress, which can quickly reduce both yield and tuber quality. Irrigation also carries substantial pumping and energy costs.

The central challenge: How can farmers apply enough water to protect yield and quality without wasting water or the energy required to pump it?

The conventional approach

Estimating Crop Water Use with a Standard Coefficient

Conventional irrigation scheduling commonly estimates crop evapotranspiration using the following relationship:

ET = ETo × Kc

ET = estimated crop evapotranspiration or crop water use

ETo = reference evapotranspiration calculated from local weather data

Kc = crop coefficient representing the crop’s growth stage and water demand

Standard crop coefficients, including those published in FAO-56 tables, are widely used. However, these coefficients are static averages based on an assumed crop-development pattern.

The actual coefficient can differ because canopy growth is influenced by planting date, variety, management, weather, soil conditions and crop stress. Fixed values can therefore result in either over-irrigation or under-irrigation.

The satellite alternative

Measuring the Crop Instead of Assuming Its Growth

The research investigated whether real-time crop coefficients could be estimated from satellite observations. The principal indicator was the Normalized Difference Vegetation Index, or NDVI, obtained from Sentinel-2 satellite imagery.

NDVI provides an indication of green vegetation and canopy development. As the potato canopy expands, the satellite records the change and allows researchers to calculate a dynamic crop coefficient known as Kc_NDVI or Kcb_NDVI.

ET = ETo × Kc_NDVI

Local weather demand multiplied by a crop coefficient derived from the crop’s observed canopy

The important difference is that the satellite coefficient responds to the crop that is actually growing in the field rather than following a predetermined growth curve.

From satellite to irrigation decision

How the System Works

Step 1

Observe the Canopy

Sentinel-2 satellite imagery is used to calculate NDVI and monitor the development and condition of the potato canopy.

Step 2

Calculate a Dynamic Coefficient

The vegetation index is converted into a crop coefficient that reflects the actual canopy rather than a standardised growth stage.

Step 3

Combine It with Weather Data

The satellite-derived coefficient is multiplied by reference evapotranspiration obtained from a local automatic weather station.

Step 4

Recommend an Irrigation Amount

The estimated crop water use can be translated into a practical irrigation recommendation for the grower.

Field validation

Testing the Method in Commercial Potato Fields

The research team monitored seven potato fields in South Africa’s Sandveld production region. The study collected information on:

• Weather conditions
• Irrigation applications
• Drainage and soil-water content
• Canopy cover and fractional interception
• Satellite vegetation indices
• Actual crop evapotranspiration

Installed equipment to measure: Weather data, Irrigation, Drainage and Soil water content 

Canopy cover and fractional interception of photosynthetically active radiation were measured approximately every 10–15 days. Eddy-covariance systems were installed in three fields to obtain direct measurements of actual evapotranspiration.

Three methods were compared

1. Standard FAO-56 crop coefficients

2. Crop coefficients derived from satellite NDVI

3. Actual ET measured through eddy covariance

Research findings

What Did the Researchers Find?

Satellite NDVI Followed Canopy Development

NDVI tracked changes in canopy cover and the fractional interception of photosynthetically active radiation effectively.

Satellite Coefficients Agreed with Measured ET

For most of the growing season, the NDVI-derived crop coefficients correlated closely with coefficients calculated from eddy-covariance measurements.

Early-Season Accuracy Was Weaker

Before the canopy completely covered the ground, evaporation from exposed soil represented an important part of total water loss. Canopy NDVI alone could not capture this component as accurately.

Dynamic Coefficients Reflected Actual Field Conditions

Unlike fixed coefficients, the satellite values responded to differences in canopy growth, crop condition and field management.

From research to farm

Testing GeoIrrigation at Commercial Scale

The concept was developed further through collaboration with GeoTerraImage. A GeoIrrigation service based on the Kcb_NDVI approach was tested on approximately 300 hectares of commercial potato fields during the 2025/26 production season.

The service converted satellite observations and local weather data into daily irrigation recommendations, which were delivered to participating growers by email. The early commercial-scale results were described as promising.

Practical significance

Satellite information becomes valuable to farmers when it is converted into a clear operational recommendation: how much water should be applied, and when?

Potential Benefits for Growers

More efficient water use: Irrigation can be matched more closely to actual crop demand.

Lower pumping costs: Avoiding unnecessary irrigation can reduce energy consumption.

Protection of yield and quality: Improved timing may help prevent damaging periods of water stress.

Field-specific management: Recommendations reflect differences between individual fields rather than relying entirely on regional averages.

Remaining Limitations

Although the results are encouraging, the system still faces several technical challenges.

Cloud cover: Persistent cloud can delay the acquisition of usable satellite imagery.

Early crop development: NDVI is less reliable when the canopy is incomplete and soil evaporation forms a large share of total water loss.

Local validation: Relationships between NDVI and crop coefficients may require calibration for different crops, regions and production systems.

A Practical Step Towards Precision Irrigation

Satellite-derived crop coefficients provide a practical opportunity to replace generalised assumptions with field-specific information on actual canopy development.

When combined with local weather-station data, these coefficients can help growers match irrigation more closely to crop water demand. This could conserve water, lower pumping costs and support more sustainable potato production in water-scarce environments.

Continued refinement of early-season estimates, improved management of cloud-related data gaps and wider commercial testing will determine how rapidly services such as GeoIrrigation can be adopted across South Africa and comparable production regions.

Could satellite-based irrigation recommendations become a standard tool for farmers in water-scarce regions?