The Canadian Prairie Soil Spectral Library
The need to sustainably meet the growing food demand under a changing climate will require improved soil management. There is also a rapid shift in agriculture towards a data-driven approach to enhance decision-making for enhanced productivity, sustainability and profitability. The Canadian prairies accounts for more than 80% of the arable land in Canada, so effective soil management in this region has national implications. However, the lack of critical soil information at precise temporal and spatial scales is a major barrier, due in part to cost.
The application of soil spectroscopy has emerged as a valuable solution. This technology enables rapid and comprehensive soil analysis through advanced modeling techniques with soil spectral libraries (SSLs). These SSLs combine soil spectra with conventionally measured soil information, creating a valuable resource for researchers and practitioners. The Canadian Prairie Soil Spectral Library (CPSSL) is therefore being developed through a collaborative effort with a multidisciplinary team to fill these critical gaps in soil information.
The CPSSL will be the first regional SSL in Canada and tailored to the unique characteristics of the Canadian prairies. The main objectives of the CPSSL are:
- acquire and harmonize legacy soil data for research, monitoring and assessment in the Canadian prairies
- establish the CPSSL with soil property data, spectral data, and metadata (e.g., location, spatial coordinates, sampling year and depth, land cover)
- develop and validate prediction models using machine learning for low-cost and efficient assessment of soil properties in the Canadian prairies.
The CPSSL employs visible and near-infrared reflectance (Vis-NIR) spectroscopy and leverages legacy soil data spanning diverse geographic regions, depths, annual and perennial cropping systems, and sampling years. It is being constructed through benchtop spectroscopic analysis of air-dried and sieved soils (< 2 mm) within the Vis-NIR spectral range (350 – 2500 nm). The library currently contains 8396 soil spectra with associated soil property data and metadata. Associated soil data include soil organic carbon, total carbon, inorganic carbon, total nitrogen, pH, electrical conductivity, sand, silt, clay, and bulk density.
The establishment of the CPSSL signifies a significant step towards addressing the scarcity of critical soil information in the Canadian prairies. This initiative promises to enhance our understanding of soil resources, foster sustainable soil management practices, and contribute to both scientific advancement and societal well-being.
For more information about the CPSSL, please contact Dr. Mervin St. Luce (mervin.stluce@agr.gc.ca)