Development of field mobile soil nitrate sensor technology to facilitate precision fertilizer management

Precision nitrogen (N) fertilizer management has the potential to improve N fertilizer use efficiency, simultaneously reducing the cost of inputs for farmers and the off-site environmental impact of crop production. Although technology is available to spatially vary sidedress N fertilizer applicatio...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Precision agriculture 2019-02, Vol.20 (1), p.40-55
Hauptverfasser: Rogovska, Natalia, Laird, David A., Chiou, Chien-Ping, Bond, Leonard J.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Precision nitrogen (N) fertilizer management has the potential to improve N fertilizer use efficiency, simultaneously reducing the cost of inputs for farmers and the off-site environmental impact of crop production. Although technology is available to spatially vary sidedress N fertilizer application rates within fields, sensor technology capable of measuring soil nitrate (NO 3 − ) levels in-real-time and on-the-go with sufficient accuracy to facilitate precision application of N fertilizers is lacking. The potential of Diamond-Attenuated Total internal Reflectance (D-ATR) Fourier Transform Infrared (FTIR) spectroscopy was evaluated as a soil NO 3 −  sensor. Two independent datasets were tested; (1) the field dataset consisted of 124 GPS registered soil samples collected from four agricultural fields; and (2) the laboratory dataset consisted of five different soils spiked with various amounts of KNO 3 (135 samples) and incubated in the laboratory. Spectra were collected using an Agilent 4100 Exoscan FTIR spectrometer equipped with a D-ATR cell and analyzed using partial least squares regression. Calibration R 2  values (D-ATR-FTIR predicted versus independently measured soil NO 3 −  concentrations) for the field and laboratory datasets were 0.83 and 0.90 (RMSE = 8.3 and 8.8 mg kg −1 ), respectively; and robust “leave one field/soil out” cross validation tests yielded R 2 values for the field and laboratory datasets of 0.65 and 0.83 (RMSE = 12.5 and 13.3 mg kg −1 ), respectively. The study demonstrates the potential of using D-ATR-FTIR spectroscopy for rapid field-mobile determination of soil NO 3 −  concentrations.
ISSN:1385-2256
1573-1618
DOI:10.1007/s11119-018-9579-0