Digital mapping of soil properties: A review

Detailed and accurate spatial soil information is important to address the issue of land degradation, climate change, biodiversity, and food security. Conventional soil mapping techniques are expensive and time consuming. So, rapid, reliable and economical techniques are required to map the soil properties over the landscape. Digital soil mapping offers a ray of light in this regard which can produce digital soil map using legacy soil data, spectral data and also the environmental covariates. In this paper, we provide a comprehensive overview about the digital mapping of soil properties. We have described systematically the evolution of digital soil mapping (DSM), the framework of DSM, input parameters (legacy soil data, environmental covariates) to DSM, spatial modeling approaches including deep and machine learning based approaches and their applications to map and model the soil properties and their prediction. DSM approaches would provide a better insight to the decision makers and farm managers to design sustainable land management practices in various situations.