Detecting soil compaction through multispectral UAV images – A spatial analysis of traffic effects and compaction patterns in arable land

Soil compaction induced by mechanical stress of heavy machinery and intensive field traffic can reduce crop yields and is one of the main threats to soil functions. Even though soil compaction is one of the most serious types of soil degradation worldwide, its detection on a large scale remains challenging. Current methods of soil compaction detection are limited to selective samplings and laboratory analyses, which are both labor-intensive.

This study examines whether field traffic effects and soil compaction can be detected non-invasively on field scale using multispectral UAV (unmanned aerial vehicle) images. The objective of the study is to establish a link between soil properties, crop yields and multispectral images in order to determine the spatial extent of soil compaction. Two fields in Northern Germany cultivated with winter wheat and winter barley served as study area during the years 2020-2022. To analyse the effects of soil compaction on soil functions and crop yields, the slurry application in spring was used as a reference event for anthropogenically induced soil compaction.

Multispectral images of the two study sites were recorded up to four times with UAV during the vegetation period. The images were used to calculate vegetation indices (e.g. GRVI, MGRVI) and the plant height. In addition, soil sampling and manual harvests were carried out in the ruts of the slurry tanker and in the non-trafficked areas. In the laboratory soil physical (e.g. dry bulk density, air capacity) and chemical (e.g. carbon content) properties were measured as well as the grain yield.

The results of the UAV analysis show linear patterns of low plant height and vegetation index values, that can be recognized over the entire extent of the study fields, particularly in spring. The linear patterns are attributed to the slurry application. As an example, the NDVI in March 2021 shows a mean value of 0.55 in the ruts of the slurry tanker, whereas the mean value in the non-trafficked field is 0.63. These findings are consistent with the laboratory analyses. The results demonstrate that the mechanical stress by the slurry tanker leads to increased dry bulk density and reduced air capacity, air conductivity and saturated hydraulic conductivity in the trafficked areas compared to the non-trafficked inner field. In addition, the grain yield in the ruts of the slurry tanker is on average about 13 % lower than in the non-trafficked field.

However, sampling time and weather conditions have a strong effect on the analysis. In 2022, for instance, weather extremes occurred with low precipitation, which lower the correlation between UAV data and soil properties and made the analysis more complex. Nevertheless, this study shows that UAV imagery is a reliable tool to detect soil compaction at a continuous spatial scale for individual fields.