Reactive nitrogen (Nr) is a term used to describe non-nitrogen gas (non-N2) forms of nitrogen (N) in the biosphere. It causes major pollution problems when it occurs in excess, and it has many sources, including fertilizers used in production agriculture. Currently there is no on-the-go soil nitrate sensor that could guide the application of the optimal amount of fertilizer, which often varies significantly within a field. We report for the first time nitrate-in-soil measurements performed on moving soil samples at concentration levels relevant for fertilizer application. An infrared emission technique called transient infrared spectroscopy (TIRS) was tested on soil samples spiked with different nitrate concentrations in the parts-per-million range and moving at a velocity of 2.6 m/s (5.8 miles per hour) in the laboratory. The TIRS Fourier transform infrared (FT-IR) spectra were modeled by partial least squares and produced a standard error of cross-validation (SECV) of 6.3 parts per million (ppm) N and an R 2 of 0.938 for 512-scan spectra. These results are compared to those using fewer TIRS scans and to those from photoacoustic spectroscopy (PAS) and diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) measurements on stationary samples. TIRS 128-, 32-, and 8-scan spectra yielded SECVs of 11.2, 11.4, and 18.4 ppm N and R 2 values of 0.800, 0.831, and 0.583, respectively. The PAS and DRIFTS measurements produced SECVs of 12.4 and 9.0 ppm N and R 2 values of 0.766 and 0.876, respectively. Reactive nitrogen (Nr) is a term used to describe non-nitrogen gas (non-N 2 ) forms of nitrogen (N) in the biosphere. It causes major pollution problems when it occurs in excess, and it has many sources, including fertilizers used in production agriculture. Currently there is no on-the-go soil nitrate sensor that could guide the application of the optimal amount of fertilizer, which often varies significantly within a field. We report for the first time nitrate-in-soil measurements performed on moving soil samples at concentration levels relevant for fertilizer application. An infrared emission technique called transient infrared spectroscopy (TIRS) was tested on soil samples spiked with different nitrate concentrations in the parts-per-million range and moving at a velocity of 2.6 m/s (5.8 miles per hour) in the laboratory. The TIRS Fourier transform infrared (FT-IR) spectra were modeled by partial least squares and produced a standard error of cross-validation (SECV) of 6.3 parts per million (ppm) N and an R 2 of 0.938 for 512-scan spectra. These results are compared to those using fewer TIRS scans and to those from photoacoustic spectroscopy (PAS) and diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) measurements on stationary samples. TIRS 128-, 32-, and 8-scan spectra yielded SECVs of 11.2, 11.4, and 18.4 ppm N and R 2 values of 0.800, 0.831, and 0.583, respectively. The PAS and DRIFTS measurements produced SECVs of 12.4 and 9.0 ppm N and R 2 values of 0.766 and 0.876, respective...
