Serva, Lorenzo and Segato, Severino and Berzaghi, Paolo and Cozzi, Giulio (2026) Dataset including 99 spectra collected using four NIR instruments, together with butter quality parameters, for the evaluation of NIR performance. [Data Collection]
Collection description
Data comprises 99 spectra of butters acquired with four (one benchtop and three handheld) instruments. Butter samples were collected from October 2023 to September 2024 following a monthly repeated-measures sampling protocol, and classified into four groups according to the production system and milk origin: - Lowland (n = 35): monthly samples from industrial dairy plants processing milk from intensive dairy herds located in the Po Valley and fed maize silage-based total mixed rations (TMR); - Mountain (n = 47): monthly samples from industrial dairy plants located in mountain areas processing milk from dairy herds fed hay-based TMR; - Alpine (n = 10): summer samples from small-scale dairy plants processing milk from transhumant dairy herds grazing alpine pastures; - Commercial (n = 7): butter samples purchased from retail outlets with no information available on milk origin or production season. The spectral range used for the calibration transfer procedure was 1102- 1600 nm, with spectral variables arranged at 2 nm intervals.
| DOI: |  10.25430/researchdata.cab.unipd.it.00001890 |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Keywords: | NIR spectra, butter quality, butter origin | |||||||||||||||
| Subjects: | Life Sciences > Applied Life Sciences, Biotechnology and Molecular and Biosystems engineering: Applied plant and animal sciences; forestry; food sciences; applied biotechnology; environmental, and marine biotechnology; applied bioengineering; biomass, biofuels; biohazard > Applied animal sciences (including animal breeding, veterinary sciences, animal husbandry, animal welfare, aquaculture, fisheries, insect gene drive) Physical:Sciences and Engineering > Physical and Analytical Chemical Sciences: Analytical chemistry, chemical theory, physical chemistry/chemical physics > Spectroscopic and spectrometric techniques |
|||||||||||||||
| Department: | Departments > Dipartimento di Medicina animale, produzioni e salute (MAPS) | |||||||||||||||
| Depositing User: | Lorenzo Serva | |||||||||||||||
| Date Deposited: | 28 Jul 2026 07:38 | |||||||||||||||
| Last Modified: | 28 Jul 2026 07:38 | |||||||||||||||
| Creators/Authors: |
|
|||||||||||||||
| Type of data: | Database | |||||||||||||||
| Contributors: |
|
|||||||||||||||
| Collection period: |
|
|||||||||||||||
| Geographic coverage: | North Italy | |||||||||||||||
| Data collection method: | Ninety-nine (n = 99) butter samples were collected from October 2023 to September 2024 following a monthly repeated-measures sampling protocol and classified into four groups according to the production system and milk origin: - Lowland (n = 35): monthly samples from industrial dairy plants processing milk from intensive dairy herds located in the Po Valley and fed maize silage-based total mixed rations (TMR); - Mountain (n = 47): monthly samples from industrial dairy plants located in mountain areas processing milk from dairy herds fed hay-based TMR; - Alpine (n = 10): summer samples from small-scale dairy plants processing milk from transhumant dairy herds grazing alpine pastures; - Commercial (n = 7): butter samples purchased from retail outlets with no information available on milk origin or production season. After sample collection, they were stored at –20 °C until NIR and wet-chemistry analysis. Butter samples were analysed using four NIR instruments: one benchtop NIR spectrometer, the FOSS DS2500 (FOSS Analytical A/S, Hillerød, Denmark; hereafter DS25), used as the master instrument, and three handheld NIR spectrometers used as target instruments. These latter were: Aurora (Grainit S.r.l., Padova, Italy), JDSU (JDS Uni-phase Corporation, Milpitas, CA, USA), and ISC (Integrated Spectronics Pty Ltd., Australia). For each sample, paired spectra were available from both instruments and were acquired in absorbance mode under comparable analytical conditions. The spectral matrices were aligned by sample identifier, so that each row of the master matrix corresponded to the same sample in the target matrix and in the reference chemical dataset. Spectral ranges were 1100-2500, 0.5 nm for DS25, 950-1650, 2 nm for Aurora, 910-1670, 2 nm for JDSU, and 1102-1600, 2 nm for ISC, respectively. The common spectral range used for the calibration transfer procedure was 1102-1600 nm, with spectral variables arranged at 2 nm intervals. Before spectral acquisition, each sample was gently stirred and placed in a round quartz-window cup. The samples were maintained at 4 °C to minimise heating effects during scanning, which could arise from exposure to room temperature and lamp irradiation. To reduce potential bias associated with the acquisition sequence, the order of instrument use was randomly varied across samples. | |||||||||||||||
| Resource language: | English | |||||||||||||||
| Metadata language: | English | |||||||||||||||
| Publisher: | Research Data Unipd | |||||||||||||||
| Date: | 27 July 2026 | |||||||||||||||
| Copyright holders: | The Author | |||||||||||||||
| URI: | https://researchdata.cab.unipd.it/id/eprint/1890 |
orcid.org/0000-0003-0660-9637
