Digital Soil Mapping: High Resolution Maps for Modern Land Management Decisions

Science Center Objects

The field of digital soil mapping has bridged the classic theories of soil science into the modern computing age to produce high resolution maps. This body of work utilizes classic soil factorial theory (soil = f[climate, organisms, relief (topography), parent material, time] + ɛ, or ‘clorpt’). The clorpt framework has been approximated using various environmental spatial data layers to predict soil types and properties between field observations, which are expensive and time consuming. 

Background & Importance

Soil profile in an arid rangeland in New Mexico

Several feet of a soil profile in an arid rangeland in New Mexico. (Credit: Mike Duniway, USGS. Public domain.)

In the age of big data and complex modeling in science, quantitatively sound, high resolution soil maps are more important than ever. Soils play an important role as an interface in mediating water and nutrient movement in ecosystems that can vary greatly across different soil types. This spatial variation in soils can mean that different areas may be more or less vulnerable to human disturbance and climate change while still maintaining ecosystem services. Soils are at the heart of the ‘potential’ of an area to produce the foundational energy of photosynthetic vegetative production, maintain surface water supply and quality, and keep particulates and dust from threatening air quality. Understanding how soils of differing potential and vulnerability are distributed through a landscape with maps can help a multitude of researchers, land owners/managers, and policy makers to make informed decisions.

General Methods

Digital soil mapping employs a vast set of environmental mapping layers with exhaustive spatial extent to help predict soil types and properties between field observations of soils. Environmental mapping layers include topographic characteristics, satellite imagery to map vegetation and sometimes geology, climate layers produced from weather stations, other remotely sensed data like gamma radiometrics, and many other useful layers depending on the map type and size. Locations where soil field observations exist are related to the environmental layers with machine learning or artificial intelligence algorithms (e.g. random forests) to help predict soil type in locations that scientists haven’t visited, but that are similar in environmental condition to places where soils have been sampled. This process is repeatable, easily updated and provides robust measures of map accuracy (cross validation) and prediction uncertainty at every pixel (often hundreds of millions of pixels in a map.

Important Results

Recent developments include new 100-meter (328’) resolution soil property and class maps of the lower 48 US states (Ramcharan et al. 2018). This builds upon the work by Hengl et al. (2017) where the entire globe was mapped at 250-meter (820’) resolution. The POLARIS project has predicted soil series across the lower 48 states at 30-meter (98’) resolution and is currently using those initial maps to make more interpretable soil property maps at 30 meter resolution for the country (Chaney et al. 2016). Projects here at the USGS, Southwest Biological Science Center have started to utilize these types of datasets for helping to manage oil and gas well-pad reclamation (Nauman and Duniway 2016, Nauman et al. 2017), and Colorado River salinity management.

References Cited

Chaney, N.W., Wood, E.F., McBratney, A.B., Hempel, J.W., Nauman, T.W., Brungard, C.W., and Odgers, N.P., 2016, POLARIS: A 30-meter probabilistic soil series map of the contiguous United States: Geoderma, v. 274, p. 54-67.

Hengl, T., de Jesus, J.M., Heuvelink, G.B.M., Gonzalez, M.R., Kilibarda, M., Blagotić, A., Shangguan, W., Wright, M.N., Geng, X., and Bauer-Marschallinger, B., 2017, SoilGrids250m: Global gridded soil information based on machine learning: PloS one, v. 12, no. 2, p. e0169748

Nauman, T.W., and Duniway, M.C., 2016, The Automated Reference Toolset: A Soil-Geomorphic Ecological Potential Matching Algorithm: Soil Science Society of America Journal, v. 80, no. 5, p. 1317-1328.

Nauman, T.W., Duniway, M.C., Villarreal, M.L., and Poitras, T.B., 2017, Disturbance automated reference toolset (DART): Assessing patterns in ecological recovery from energy development on the Colorado Plateau: Science of The Total Environment.

Ramcharan, A., Hengl, T., Nauman, T., Brungard, C., Waltman, S., Wills, S., and Thompson, J., 2018, Soil Property and Class Maps of the Conterminous United States at 100-Meter Spatial Resolution: Soil Science Society of America Journal