Intelligence for Soil (I4S)

On many farms, arable land is farmed conventionally on a standardised basis. However, this does not take into account the sometimes considerable variability of soil properties and thus the fertiliser requirements within the field. This often leads to inefficient fertilisation: Some areas receive too much fertiliser, which wastes resources and leads to environmental pollution, while others remain under-fertilised, which means that the yield potential is not exploited.

Site-specific management can adapt fertilisation to these different requirements. To do this, it is necessary to record the soil variability in high resolution. Laboratory analyses are too expensive and time-consuming for this. Soil sensors offer a quick and cost-effective solution. Although these technologies are available, they are still rarely used in agriculture. In order to increase the acceptance of precision agriculture in practice, the high-resolution sensor data must be integrated into an easy-to-use decision support system that is customised to the needs of farmers.

The aim of the ‘I4S - Intelligence for Soil’ project is therefore to develop a system that provides site-specific recommendations for fertilisation and other management measures in order to maintain soil fertility and reduce environmental pollution.

As part of I4S, a multi-sensor platform, the so-called RapidMapper, was developed and tested for the high-resolution mapping of topsoil. It is the only sensor platform in the world that is equipped with the four sensor technologies pH electrodes, electrical conductivity, near-infrared and gamma spectrometry (see image).

With this combination, a range of soil properties can be reliably recorded, including soil texture, humus content, pH value and macronutrients. The quality of the derived soil property maps can be further improved in the future by merging the data from different sensors.

In addition to these established sensors, the project is also researching new technologies for soil mapping such as X-ray fluorescence, laser-induced plasma spectroscopy and Raman spectroscopy. These sensors offer the possibility of collecting even more precise data and further improving the predictability of soil properties.

Another innovation of the project is the development of a second multi-sensor platform, the so-called RapidProfiler, which will make it possible to scan soil profiles down to a depth of one metre. This diverse data can be used to generate 3D soil maps of the entire root zone.

To better utilise the sensor data, soil-plant models have been integrated into I4S, which simulate the dynamics of nutrients and water over the course of the vegetation period. Based on the sensor data, static and dynamic models, a decision support system was developed to generate soil property maps and high-resolution fertiliser application maps.

About the Project:

Goal and approach

Most important findings

Outlook

People and Partners

Project Leaders

Partner Organizations

Contacts

News

Publications