Soil classifications are organized systems for sorting the ground beneath your feet into named categories based on measurable properties like texture, mineral content, color, drainage, and the processes that formed it. No single universal system exists. Instead, several major frameworks operate in parallel around the world, each built for different purposes and user groups. The two most widely referenced scientific systems are the USDA Soil Taxonomy, used primarily in the United States, and the World Reference Base for Soil Resources (WRB), which serves as an international standard. But engineers, farmers, foresters, and even indigenous communities have their own classification schemes, and these sometimes agree with the scientific systems and sometimes diverge in revealing ways.
The Two Dominant Scientific Systems
USDA Soil Taxonomy organizes soils into a hierarchy of six levels, from broad “orders” at the top down to individual “series” at the bottom. There are 12 soil orders in the current system, distinguished by the presence or absence of specific diagnostic horizons (layers within a soil profile that have measurable characteristics like a certain clay content, organic matter accumulation, or chemical composition). The names are built from Latin and Greek roots, so once you learn the pieces, names like “Mollisol” (soft soil) or “Aridisol” (dry soil) give you a rough idea of what you are looking at.
The WRB, maintained by the International Union of Soil Sciences, takes a somewhat different approach. Rather than a deep hierarchy, it uses 32 “Reference Soil Groups” as its backbone. These groups are then modified with qualifiers that describe additional features. The WRB was designed to facilitate communication across national borders, so it deliberately tries to bridge the vocabulary gaps between countries that developed their own systems independently.
Russia has its own long-standing classification tradition, which predates both modern systems. When researchers have tried to quantitatively compare how these three systems handle the same soils, the correspondence is often high but not perfect. A study correlating Russian, American, and WRB classifications of dark grassland soils found that most Russian soil types mapped closely to WRB Reference Groups and to the Great Groups of USDA Soil Taxonomy, with one notable exception: the Kastanozems (chestnut-colored grassland soils) did not align well from a geographic standpoint, even though the measured properties were similar.1Catena. “Black soils” in the Russian Soil Classification system, the US Soil Taxonomy and the WRB That kind of mismatch matters because a soil classified identically in two systems might actually occur in different landscapes, complicating global soil maps.
Why One Soil Can Land in Different Categories
The systems do not always slice the world the same way because they emphasize different diagnostic criteria. A good example involves Vertisols, the clay-rich soils that swell dramatically when wet and crack when dry. In USDA Soil Taxonomy, Vertisols are defined primarily by their shrink-swell behavior, which makes them quite distinct from Mollisols (the fertile, dark-colored grassland soils). But the WRB’s definition of its Vertisol Reference Group captures some soils that, when measured quantitatively, cluster more closely with the Mollisol-like group.2Catena. “Black soils” in the Russian Soil Classification system, the US Soil Taxonomy and the WRB This is not an error in either system; it reflects genuinely different ideas about which soil properties should carry the most weight when drawing category boundaries.
The same tension plays out with volcanic-ash soils. USDA Soil Taxonomy places them in the order Andisols, defined by the presence of short-range-order minerals like allophane, which form from rapidly weathered volcanic glass. These minerals give Andisols unusual properties: they fix phosphorus aggressively, making it less available to plants, and their behavior changes depending on the composition of the parent ash. Indonesian Andisols along the volcanic arc of Java, for instance, show increasing allophane content from east to west, which directly raises their capacity to lock up phosphorus.3Geoderma. Surface reactivity of Andisols on volcanic ash along the Sunda arc crossing Java Island, Indonesia Because these properties are so distinctive, Andisols occupy their own order in Soil Taxonomy, but they can superficially resemble dark Mollisols in color and organic matter content. The quantitative analysis of “black soils” recommended excluding both Andisols and Vertisols from the dark-soil cluster because their management needs are so different.4Catena. “Black soils” in the Russian Soil Classification system, the US Soil Taxonomy and the WRB
Engineering Classifications Serve a Completely Different Purpose
If you are building a road, a runway, or a foundation, you do not care much about how a soil formed over millennia. You care about how it will behave under a load, whether it will drain, and whether it will shrink or swell. Engineering soil classifications were built to answer exactly those questions, and they sort soils by grain size and plasticity rather than by pedogenic horizons.
The most widely known engineering system is the Unified Soil Classification System (USCS), which dates back to Arthur Casagrande’s work in the 1940s. It uses a two-letter code (GW for well-graded gravel, CH for high-plasticity clay, and so on) based on sieve analysis and Atterberg limits, which measure the moisture content at which a fine-grained soil transitions from solid to plastic to liquid behavior. The American Association of State Highway and Transportation Officials (AASHTO) system serves a similar role but is tuned for road construction.
A critical review of fine-grained soil classification charts found that while the Casagrande-USCS approach is by far the most widely used, it is not necessarily the most accurate. The review examined six major systems and their variants, all built around Atterberg limits, and concluded that only one of the newer proposals demonstrated strong predictive capacity with well-founded criteria.5ScienceDirect. Review and critical examination of fine-grained soil classification systems based on plasticity The USCS persists largely because of institutional momentum and the enormous body of engineering data already keyed to it, even though its boundaries between soil groups are somewhat arbitrary.
For the average person, the practical takeaway is that a soil report from a geotechnical engineer and a soil report from a soil scientist may describe the same patch of ground in completely incompatible language. The engineer’s “CL” (low-plasticity clay) and the soil scientist’s “Vertisol” might even refer to different properties of the same material.
Soils That Swell, Crack, and Damage Structures
Vertisols deserve special attention because they cause billions of dollars in property damage worldwide. These soils are dominated by smectite clays that absorb water and expand, then shrink and crack as they dry. The cycle produces distinctive features: deep, wide cracks in the dry season, slickensides (polished surfaces where soil masses slide against each other), and a self-churning process that mixes material from different depths.
In Canada, these soils fall under the Vertisolic order and develop in fine-textured parent materials with high shrink-swell potential. They create problems for both agriculture and engineering because the constant volume changes can heave foundations, buckle roads, and make tillage unpredictable.6Canadian Journal of Soil Science. Vertisolic soils of Canada: Genesis, distribution, and classification If you have ever seen a house with stair-step cracks running through the brick, there is a reasonable chance it sits on expansive clay.
Deeply Weathered Tropical Soils
At the opposite extreme from the young, mineral-rich volcanic soils are the Oxisols (called Ferralsols in the WRB and Latossolos in Brazil), the most intensely weathered soils on Earth. They form under prolonged tropical conditions where heavy rainfall leaches out most nutrients and leaves behind a thick, uniform profile dominated by iron and aluminum oxides. They are typically deep, well-drained, and nutrient-poor.
A study of Brazilian Latossolos across different biomes found that the reddest suborder (Latossolos Vermelhos) represented the greatest degree of weathering, with the highest iron and aluminum oxyhydroxide content, the most developed microaggregates, and the lowest levels of available phosphorus. The yellowish suborder (Latossolos Amarelos) had more kaolinite clay and sand, along with higher aluminum saturation.7ScienceDirect. Brazilian Latossolos (Ferralsols, Oxisols) from different biomes These distinctions matter enormously for tropical agriculture. Farming on Oxisols almost always requires heavy liming and fertilization to compensate for the soil’s natural poverty, and the specific subtype tells you which amendments you will need most.
Permafrost Soils in a Warming World
Soils underlain by permafrost pose their own classification challenges because their defining feature, permanently frozen ground, drives a set of processes found nowhere else. Cryoturbation, the physical churning caused by repeated freeze-thaw cycles, distorts soil horizons into irregular, folded shapes that look nothing like the neat horizontal layers in a textbook profile. Cryoturbated horizons are recognized as diagnostic features in the soil taxonomies of Canada, the United States, Russia, and the WRB.8Geoderma. Recognition of cryoturbation for classifying permafrost-affected soils
In Canada, the Cryosolic order includes both mineral and organic soils that have permafrost within the top one to two meters. The order is split into three great groups: Turbic (showing clear signs of cryoturbation), Static (permafrost present but without obvious churning), and Organic (peat soils with permafrost).9Canadian Journal of Soil Science. Cryosolic soils of Canada: Genesis, distribution, and classification As the climate warms and permafrost thaws, some of these soils will lose the very feature that defines them, creating a slow-motion reclassification event across the Arctic.
Human-Made Soils
Cities, mines, and centuries of intensive agriculture have created soils that do not fit neatly into systems designed for natural landscapes. The WRB addresses this with two Reference Groups: Anthrosols, formed by long-term human additions of organic or inorganic materials to a natural soil (think of the deep, dark garden soils in European cities that have been amended for centuries), and Technosols, formed in materials deposited by human activity such as mine spoils and urban fill.10International Encyclopedia of Geography. Soils of Urban and Human‐Impacted Landscapes
Canada has proposed adding an Anthroposolic order to its national system to handle these soils. The proposed order covers soils where one or more natural horizons have been removed, replaced, added to, or significantly modified by human activity. Three great groups are defined by the presence of human-made artifacts and organic carbon content, with eight subgroups based on factors like the thickness of disturbed horizons, material composition, and whether permafrost is present.11Canadian Journal of Soil Science. Revised proposed classification for human modified soils in Canada: Anthroposolic order The need for such categories reflects a broader reality: in many parts of the world, the most common soil a person actually encounters is one that has been reshaped by human hands.
Indigenous Soil Classification Systems
Long before pedology existed as a science, farming communities around the world developed their own soil classification systems based on generations of hands-on experience. These systems tend to emphasize properties that matter directly for cultivation, particularly texture and color, rather than the subsurface diagnostic horizons that dominate scientific taxonomy.
Research in the Andes of southern Peru documented a hierarchical indigenous classification with up to four categorical levels and roughly 50 names for different soil and earth materials. The system centered on texture, and laboratory analysis showed a fairly close match between the indigenous texture classes and those used in USDA soil classification.12Soil Science Society of America Journal. Indigenous Knowledge and Classification of Soils in the Andes of Southern Peru In a Mesoamerican community, spatial analysis using GIS found that local soil maps and technical soil maps (based on USDA Soil Taxonomy) agreed on roughly three-quarters of the mapped area at broad classification levels, dropping to about 60% agreement at finer levels.13Geoderma. Local soil classification and comparison of indigenous and technical soil maps in a Mesoamerican community using spatial analysis
In Madagascar, farmers classified soils mainly by topsoil color and texture. Darker soils, called “Mainty” (black) and “Mena” (red), were considered the most fertile, and this perception lined up with scientific soil assessments. Interestingly, though, the classification systems varied between villages even within the same region, reflecting differences in local environments, ethnic origin, and traditional livelihoods.14Geoderma Regional. Ethnopedological knowledge and soil classification in SW Madagascar The takeaway from this body of work is that indigenous systems are not primitive approximations of scientific classification. They capture genuinely useful information, sometimes information that scientific surveys miss because they focus on different diagnostic criteria.
Classifying Ancient Soils
Paleosols, soils that formed on landscapes of the geologic past, present a unique classification problem. They come in three varieties: buried (covered by younger sediments), exhumed (once buried, now re-exposed at the surface), and relict (formed under past conditions but never buried).15CATENA. Paleosol classification: Problems and solutions The challenge is that burial and time alter a soil’s original properties. Organic matter decomposes, colors change, and minerals transform, which means applying modern classification keys to a paleosol can produce misleading results.
Researchers have proposed property-based classification systems specifically for paleosols, linked to the genetic processes that formed them rather than to the modern diagnostic criteria that may no longer apply. Standardized description and sampling protocols are needed because paleosols serve as archives of past climate, ecosystems, and environments, and their usefulness depends on consistent, comparable data collection across studies.16Earth-Science Reviews. A review and field guide for the standardized description and sampling of paleosols A paleosol preserved in a sedimentary rock sequence can tell you whether a given location was a tropical forest, a grassland, or a desert millions of years ago, but only if it is described and classified with enough rigor to support that interpretation.
Forest and Ecological Site Classification
Soil classification also feeds into broader ecological land classification systems, which combine soil data with climate and vegetation to describe what a given site can support. In Belgian forest ecology, for example, sites are classified along gradients of nutrient supply and moisture. The nutrient regime runs from extremely poor (hyper-oligotrophic) through balanced (mesotrophic, where tree growth faces virtually no nutrient limitation) to carbonated soils where excess calcium creates its own problems of nutrient imbalance. The moisture gradient spans from waterlogged sites with oxygen-deprived roots all the way to excessively dry conditions.17ScienceDirect. Prediction of forest nutrient and moisture regimes from understory vegetation with random forest classification models
These ecological classification systems sit on top of soil taxonomy rather than replacing it. A forester choosing which tree species to plant needs to know not just “this is an Alfisol” but where on the nutrient and moisture gradients the specific site falls. The soil classification gives you the raw material; the ecological classification tells you what to do with it.
Digital Soil Mapping and the Future of Classification
Traditional soil mapping involves walking the landscape, augering holes, describing profiles, and drawing boundaries by hand. It is slow, expensive, and in many parts of the world, still incomplete. Digital soil mapping uses satellite imagery, terrain data, and machine learning to predict soil types across landscapes where field data is sparse.
A study in arid Iran compared several machine learning approaches for predicting soil great groups and subgroups using environmental variables, including spectral data from Sentinel-2A satellite images and radar data from ALOS-PALSAR. The research evaluated how well models could capture not just surface conditions but also subsurface horizon patterns.18Advances in Space Research. Digital soil mapping for soil types using machine learning approaches at the landscape scale in the arid regions of Iran The promise of this approach is enormous: automated classification across entire regions using data that is freely available from satellites. The limitation is that soils are three-dimensional objects, and remote sensing primarily sees the surface. Predicting what lies below from what is visible above remains the central challenge.
These tools do not replace traditional classification systems; they operationalize them. The machine learning model still predicts categories defined by USDA Soil Taxonomy or the WRB. What changes is the speed and spatial coverage at which those predictions can be made, which is particularly valuable in developing regions where detailed soil surveys have never been conducted.
Why Multiple Systems Persist
Given all the confusion, you might wonder why the world has not simply agreed on one soil classification. The short answer is that different users need different things from a classification. A farmer wants to know fertility and drainage. A civil engineer wants to know bearing capacity and plasticity. A climate scientist wants to know carbon stocks and permafrost depth. A geologist studying paleosols wants to reconstruct ancient environments from degraded evidence. No single classification efficiently serves all these purposes, and attempts to build a universal system tend to produce something too complex for any one user group to apply practically.
The WRB comes closest to a global standard for scientific purposes, but even it coexists with national systems because countries have decades of soil survey data keyed to their own categories. Switching systems would mean re-mapping or at least re-labeling every soil map in the country, a task so enormous that it rarely happens. Instead, correlation tables are built to translate between systems, and researchers quantify how well categories in one system correspond to categories in another. The imperfect overlaps in those correlations are not just bureaucratic inconveniences; they reveal genuine disagreements about which soil properties matter most, disagreements that are unlikely to be fully resolved because the answer depends on what you are trying to do with the soil.