There is great value in traditional USDA style soil maps. The series concept allows you to map out regions where soil properties can be grouped together under one umbrella of management. For a land use standpoint, the technicality of the classification and how properties change across space is less important than answering one simple question- can I manage this soil the same way? In other settings, property maps could be more useful.

Both are necessary. For hydrological studies, for water retention capacity for leaching of minerals and macro nutrients, world need traditional maps. For agricultural development attribute maps are required. For improving soil we need a new type of attribute that maps amount or number per cubic centimeter of organism in soil. Soil is not simply an amalgamate of rock particles, nutrients and minerals, it is also living thing in a sense.

There is great value in traditional USDA style soil maps. The series concept allows you to map out regions where soil properties can be grouped together under one umbrella of management. For a land use standpoint, the technicality of the classification and how properties change across space is less important than answering one simple question- can I manage this soil the same way? In other settings, property maps could be more useful.

I think we - the world - need both types of maps :) . However, I work more with the soil attrbutes maps (digital soil map) ... ;)

I think that neither conventional maps, nor soil properties maps are enough to meet the demand of soil information. The methodology to create conventional soil maps is not transparent at all. The raw data is unknown and, in most cases, the accuracy is not assessed (It happen in my country, Argentina). Soil classification is also ambiguous and the real differences (distance) between soil classes are not clear. However, soil properties maps lacks of integration among them. It is not possible to interpret the soil forming processes that are taken (or have taken) place in the soil from soil properties maps, which tell us how a soil works.

I tend to think that rather needing both conventional and digital soil mapping, we need a better approach to create soil maps, considering well-documented methodology as well as soil properties interrelationships integration.

Isn't the answer in what your needs are? Pedological soil maps are useful for inferring co-varying soil properties (CECsoil varies with clay content, soil structure, plasticity index, etc.) of that soil pedogenetic soil type e.g. Vertic soils.

On the other hand all soils have a clay content it that is what is important the spatial variation in all soil types would be more useful but less useful for extrapolation.

Multi-attribute polythetic soil classes and their derived mapping units are needed when holistic interactions must be considered.  Single attribute monothetic classes faciltate focus on specfic practical problems, but may miss interactive effects.  

I think both conventional soil maps and soil attribute maps are important to understand soil. Soil map units have played major roles in agency-operated soil surveys covering regional, national, and global scales. Recently, needs for soil data are more diverse with pronounced environmental-centered drivers requesting high resolution, pixel-based soil products, which are associated with error assessment. On the other hand, traditional soil mapping explicitly incorporate pedological knowledge into the soil survey product, but have become costly and time-consuming when compared to emerging digital soil mapping (DSM) approaches, such as remote sensing, univariate and multi-variate statistical, geostatistical and diffuse reflectance spectroscopy.

It may be time to overcome reductionist soil conventional or attribute maps by integrating soilscapes with land utilization types as a complex systemic approach.

If real soils are mapped, soil properties are mapped in the process. To map real soils we need to understand the relationships better. I think understanding soilscapes help.

Hello,

as conventional soil map (with soil names) gives you information on the spatial organisation of the horizons, layers and parent material , I think that those maps permit to better understand and predict spatial organization of soil units in the landscape.

However, when dealing with certain issues (agricultutre, erosion, hydrology) it is necessary to "generate" attribute maps focusing on soil attributes really relevant regarding the purpose of the study.

So, for regions where no information are available, "generating" conventionnal soil maps is necessary but, when you need to focus on certains environmental issues, it is necessary to create "attribue maps".

Finally, both are linked because "attributes" come from the soil description and permit to generate "conventional soil maps".What is important is to well record the soil parameters (attributes)

Following up on previous answers, I think it is worthwhile to think of soil class maps as combinations of soil attributes. How they are combined and classified is a matter of purpose, but serves an important role for generalization. Simplifying, synthesizing, and/or interpreting spatial soil knowledge are really important when communicating this knowledge to audiences outside of soil science (e.g. a farmer or city planner may not find ten different soil attribute maps to be very useful). It can also be helpful to us soil scientists as it has similarities to GIS map overlays, allowing us to more easily visualize the spatial patterns of particular combinations of soil attributes.

I would also like to point out the additional utility of polygon maps. Whether generated by conventional or digital methods there are some unique analyses that can be conducted on polygons (i.e. shape and orientation), which can be useful for investigating soil-landscape relationships. Some examples of polygon spatial analysis can be found in the following texts:

Bunge, W., 1962. Theoretical Geography. First Edition. Lund Studies in Geography Series C: General and Mathematical Geography. Lund, Sweden: Gleerup.

Hole, F.D., Campbell, J.B., 1985. Soil landscape analysis. Rowman & Allanheld, Totowa.

I agree with Bradley also,

generating soil atributes map with GIS is evident, however you need to have a good "attributes" database. I am working with the WRB classification system and what I like with this system is that with the prefix and sufixes you can create attributes soil maps.

I.e if you need to map if soil are stony  (skeletic qualifier) you can extract the skeletic qualifier and create a map independently of the Reference Soil Group (full soil name).

The same for shallow soil, represented by Leptosol or Leptic qualifier for other RSG.

In my point of view, even if Leptosol are implicitly Leptic, I will add qualifier leptic in GIS to be able to make a Leptic attribute soil map. Actually some qualifier are not available for some RSG because they are redondant with the RSG name, or theorically not possible, however when dealing with GIS map it is usefull to add them.