Spatial and Temporal Field-Scale Accuracy Assessment of a Multi-Sensor Spade for In Situ Soil Diagnostics: Performance and Limitations of the Stenon FarmLab for Precision Agriculture

Real-time, in situ soil diagnostics are increasingly relevant for precision agriculture, but their efficacy under varying field and climatic conditions remains underexplored. This study assesses the 2022/23 version of the Stenon FarmLab, a multi-sensor soil analysis tool, over a 10-month period and across 1187 measurements on six fields (five cropped, one grassland) in northeast Germany. Despite the common approach of comparing a field sensor against lab results, in this paper, the FarmLab's outputs are benchmarked using various approaches, such as time series, correlation, and geostatistical analysis, to fully evaluate the temporal and spatial stability and alignment with known soil heterogeneity. While physical soil parameters such as temperature and soil texture showed robust detection accuracy, key agronomic metrics-including mineral nitrogen (Nmin), soil organic carbon (SOC), and phosphorus-exhibited poor temporal consistency and low correlation with expected spatial patterns. Measurement errors and high sensitivity to weather conditions restrict data quality, particularly under frost and drought. Spatial clustering of more temporally stable parameters (e.g., pH, soil texture) allowed for limited zone delineation. We conclude that while the FarmLab shows partial potential for on-site soil sensing, significant limitations in nutrient measurement reliability currently prevent its use in operational precision agriculture. Enhancements in sensor calibration, environmental compensation, and software are needed for broader applicability.