Principal Investigators:

Summary:

Primary References:

Webb, Rosenzweig, and Levine Global Soil Particle Size Properties

• Integrated dataset
• Design
• Source
• Additional references
• File lists

Dataset Description

INTEGRATED DATA­SET

Data­Set Citation:

Projection:

Spatial Representation:

Temporal Representation:

Data Representation:

Layers and Attributes:

Dataset Description

DESIGN

Variables:

Origin:

Geographic Reference:

Geographic Coverage: Global

Geographic Sampling:

Time Period:

Temporal Sampling:

Dataset Description

SOURCE

Source Data Citation:

Contributor:

Distributor:

Date of Production:

Lineage & Contacts:

(2) R.S. Webb
NOAA National Geophysical Data Center 

Dataset Description

ADDITIONAL REFERENCES

Bouwman, A.F., Fung, I.Y., Matthews, E.E., and John, J.G. 1991. Global model of Nitrous Oxides production from natural soils. Global Biogeochemical Cycles, submitted.

Buol, S.W., Hole, F.D., McCracken, R.J. 1973. Soil Genesis and Classification. The Iowa State University Press, Ames, Iowa.

Delworth, T.L., and Manage, S. 1988. The influence of potential evaporation of the variabilities of simulated soil wetness and climate. Journal of Climate, 1, 523­547.

FAO­UNESCO, 1971­1981. Soil Map of the World, 1:5,000,000, Volumnes II­X. UNESCO, Paris.

Hansen, J., Russell, G., Rind, D., Stone, P., Lacis, A., Lebedeff, S., Reudy, R., and Travis, L. 1983. Efficient three­dimensional global models for climate studies. Monthly Weather Review, 111, 609­662.

Henderson­Sellers, A., Wilson, M.F., Thomas, R., and Dickinson, R.E. 1986. Current Global Land­Surface Data Sets for Use in Climate­Related Studies. NCAR Technical Note NCAR/TN­272+STR.

Kellog, W.W., and Zhao, Z.C. 1988. Sensitivity of soil moisture to doubling of carbon dioxide in climate modeling experiments, I, North America. Journal of Climate, 1, 348­366.

Matthews, E. 1984. Prescription of Land­Surface Boundary Conditions in GISS GCM II: A simple method based on high­resolution vegetation data bases. NASA Technical Memorandum #86096.

Matthews, E. 1983. Global Vegetation and land use: New high­resolution data bases for climate studies. Journal of Climate and Applied Meteorology, 22, 474­487.

Petersen, G.W. Cunningham, R.L. Matelski, R.P., 1968. Available moisture within selected Pennsylvania soil series. Pennsylvania State University Agronomy Series #3, 21pp.

Rind, D. 1988. The Doubled CO2 Climate and the Sensitivity of the Modeled Hydrologic Cycle. Journal of Geophysical Research, 93 (D5), 5386­5412.

Rind, D., Goldberg, R., Hansen, J., Rosenzweig, C., and Ruedy, R. 1990. Potential evapotranspiration and the likelihood of future drought. Journal of Geophysical Research, 95 (D7), 9983­10004.

Soil Science Society of America 1987. Glossary of Soil Science Terms. Soil Science Society of America. Madison, WI.

Webb, R.S. 1990. Late Quaternary Water­Level Fluctuations in the Northeastern Unites States. Brown University Ph.D. thesis, Providence, RI.

Zobler, L. 1986. A World Soil File for Global Climate Modeling. NASA Technical Memorandum #87802. 

Dataset Description

FILE LISTS

• Dataset Files
• Reprint Files

Webb, Rosenzweig, and Levine Global Soil Particle Size Properties

• Continent Codes
• Zobler Soil Type
• Potential Storage of Water in Soil Profile
• Potential Storage of Water in Root Zone
• Soil Water Model II
• Soil Profile Thickness
• Texture-based Potential Storage of Water
• Depth of Horizon
• Proportion of Clay in Horizon
• Proportion of Sand in Horizon
• Proportion of Silt in Horizon

Continent Codes

Description:

Structure:

Series:

System Files:

Notes:

Zobler Soil Type

Description:

Structure:

Series:

System Files:

Notes:

Potential Storage of Water in Soil Profile

Description:

Structure:

Series:

System Files:

Notes:

Potential Storage of Water in Root Zone

Description:

Structure:

Series:

System Files:

Notes:

Soil Water Model II

Description:

Structure:

Series:

System Files:

Notes:

Soil Profile Thickness

Description:

Structure:

Series:

System Files:

Notes:

Texture-based Potential Storage of Water

Description:

Structure:

Series:

System Files:

Notes:

Depth of Horizon

Description:

Structure:

Series:

System Files:

Notes:

Proportion of Clay in Horizon

Description:

Structure:

Series:

System Files:

Notes:

Proportion of Sand in Horizon

Description:

Structure:

Series:

System Files:

Notes:

Proportion of Silt in Horizon

Description:

Structure:

Series:

System Files:

Notes:

Webb, Rosenzweig, and Levine Global Soil Particle Size Properties

Mark Ohrenschall, NOAA/NESDIS
National Geophysical Data Center
Boulder, CO

The following is an excerpt from documentation provided by Robert S. Webb. This selection refers to a data file containing depth and soil particle size information. Note that references to 106 entries in the data array (corresponding to Zobler soil types) is a typographical error, and the correct figure is 107.

The data has been organized as four 106x10x15 dimensioned real*4 arrays: depth, sand, silt, and clay. The first dimension (106) corresponds to the sequence number of the soil types in Zobler's (1986) World Soil Data File. The second dimension (10) corresponds to the volume numbers of the nine major continental divisions in FAO/UNESCO Soil Map of the World, Vols. 2­10 (1971­81). The third dimension (15) corresponds to the individual horizons with data for each soil type from the Morphological, Chemical and Physical Properties Appendix in each of the nine volumes of the FAO/UNESCO Soil Map of the World (1971­81). The data in the sand, silt, and clay arrays are stored as proportional values for each soil horizon. The arbitrary particle size distribution summing to 100 percent included for Histosols (entries 61­63 in first dimension of each array) should not be used. Instead, values reflecting the physical properties of organic soils and appropriate for specific research objectives should be inserted.

The data in the depth array are scaled in meters with the first value being 0m depth for each soil type and the subsequent values the contact depths of contiguous horizons. By definition the depth array contains one extra value for the third dimension corresponding to the bottom depth of the lowest horizon for each soil type. Within the data set, no soil type had more than 14 soil horizons. In cases when the number of horizons in a soil type was less than 14, we used ­1.0 values to flag the end of record of each soil type. For example, a soil type with 10 horizons has 10 data entries in the sand, silt, and clay arrays, 11 data entries for the depth array, and ­1.0 values for entries 11 ­ 15 in each array (entries 12 ­ 15 for the depth array).

Some technical notes are given regarding the 107x10x15 data array for those interested:

1) A code for ocean was added to the group of nine continent codes, thus accounting for the 10 elements of the second dimension of the data array. The data array for all soil horizons for all soil types for this continent code was zero-filled.

2) The data array was an ASCII text file with four columns of numbers, each column corresponding to one of the four variables, namely depth, sand, silt, or clay. Thus each array element was actually a line of text containing four data values for the four variables.

3) The ordering of the array elements into the (one-dimensional) data file was such that the 107 soil types vary slowest, the 10 continent codes vary faster, and the 15 soil horizons vary fastest. In other words, if an element's position in the array is given by the indices (i,j,k) where 1 <= i <= 107, 1 <= j <=10, and 1 <= k <= 15 then the position of that element in the data file is given by ((i - 1) *15 * 10) + ((j - 1) * 15) + k = ((i - 1) * 10 + j - 1) * 15 + k.

The first stage in producing the IDRISI format for the data array was to separate the data by variable (depth, sand, silt, and clay) and by horizon number (one through 15) into 60 attribute values files. Each attribute values file would be composed of feature identification codes corresponding to each of the 107 soil types for each of the 10 continent codes (explained below), with each feature I.D. being paired with a data value. The data value for each feature I.D. was read from the appropriate position in the data array (given above). In other words, the first and second dimensions of the data array were merged into a single dimension with 107 * 10 = 1070 elements, and the third and fourth dimensions (the fourth dimension is the variable) were also merged into a single dimension with 15 * 4 = 60 elements. Here the elements of the first merged dimension are "continental soil type" (the feature I.D.'s) and data value pairs , and the elements of the second merged dimension are attribute values files, named after variable and soil horizon.

The second stage in producing the IDRISI format was to create the spatial map associated with the attribute values files. This spatial map would be the feature definition file that uses the continental soil types as links between the data values and geographic locations. Since the soil types and the continental divisions are already spatially defined it only remained to produce the map of continental soil types. This was done by overlaying the map of continent codes (WRCONT) multiplied by 1000 with the map of soil types (WRZSOL) via addition¹. Both the original continent codes and the original soil types can be recovered from this map, the continent code by performing integer division by 1000, and the soil type by taking the continental soil type modulo 1000.

The final stage in producing the IDRISI format was to produce 60 separate raster grids from the 60 attribute values files and the single feature definition file. This was done by running the IDRISI module ASSIGN on the feature definition file and on each of the 60 attribute values files. The ASSIGN module creates an output grid from an input grid and an attribute values file, using the input grid (whose cells take on feature I.D.'s as values) to define the locations of the data values found in the attribute values file. The appropriate data values are taken from the attribute values file according to the feature I.D.'s paired with each data value. Thus if a cell in the input grid has a value p and the attribute values file has a feature I.D. and data value pair (p,z) then the cell with the corresponding position in the output grid will take on the value z. Note that feature I.D.'s in attribute values files must be unique, but feature I.D.'s in the feature definition file may occur multiple times.