CIDC FTP Data
Soil IDC Data on FTP

Data Access

FAO Soil Data
rule

Readme Contents

Data Set Overview
Sponsor
Original Archive
Future Updates

The Data
Characteristics
Source

The Files
Format
Name and Directory Information
Companion Software

The Science
Theoretical Basis of Data
Processing Sequence and Algorithms
Scientific Potential of Data
Validation of Data

Contacts
Points of Contact

References

rule

Data Set Overview


Climate modelers need information on the water holding capacity of global soils. Currently the best source of this information is the Soil Map of the World, which was produced by the Food and Agriculture Organization (FAO) of the United Nations Educational, Scientific, and Cultural Organization (UNESCO) in 10 volumes between 1970 and 1978. It provides the most detailed, globally consistent soil data.

Because water holding capacity is not an explicit attribute of the FAO soil map, the data on soil type, soil texture, soil depth, and average slope that the soil map does provide may be used as surrogates. The four data sets described herein were derived, by various researchers, from the FAO soil data. For climate modelers, a 1 degree by 1 degree grid of latitude and longitude has been deemed adequate.

Sponsor
The production and distribution of this data set are being funded by NASA's Earth Science enterprise program. The data are not copyrighted; however, we request that when you publish data or results using these data please acknowledge as follows:

The authors wish to thank the Distributed Active Archive Center (Code 902 .2) at Goddard Space Flight Center, Greenbelt, MD, 20771, for producing the data in its present format and distributing them. The original data sets where produced by:

  • Soil type and soil texture were constructed by Zobler (1986).
  • Soil profile depth data set was constructed by Webb et al. (1993).
  • Slope data were originally derived from the FAO Soil Map of the World in a 1 degree grid (GLOBTEX), version 1.0, by the Science and Applications Branch, EROS Data Center, Sioux Falls, South Dakota.
  • Dr. R.D. Koster performed the analyses necessary to assign parameter values to the soil map texture classes.

Goddard's share in these activities was sponsored by NASA's Earth Science enterprise.

Original Archive
The original source maps are the FAO Soil Map of the World. The Earth Sciences and Resource Institute (ESRI) digitized the data under contract to the United Nations Environment Program (UNEP) and the FAO in 1984. The EROS Data Center constructed the data sets that were later used to derive the global array of average slope.

The soil type file was acquired from Goddard Institute of Space Science. The soil texture, slope and depth files were acquired from the ISLSCP Initiative I CD set.

Future Updates
An international effort to develop a replacement, the Soil and Terrain (SOTER) digital data base of the world, is under development by the International Society of Soil Science, the International Soil Reference and Information Center, the FAO, and the UNEP.

The Data

Characteristics

Source
Digital data for the FAO Soil Map of the World are available from the Land and Water Development Division, FAO, in Rome, Italy.

A) SOIL TYPE. The soil type data file was derived from the highest level of the FAO soil units and is based on the work of Zobler (1986).

B) SOIL TEXTURE. The soil texture data file is based on the work of Zobler (1986) and uses the indices listed in the table below to identify the texture of the dominant soil type within each 1 degree x 1 degree grid square. The original FAO data provided, for the dominant soil type in a soil unit, the designation "coarse", "medium", "fine", or a combination of these based on the relative amounts of clay, silt, and sand present in the top 30 cm of soil.

Also listed in the table are some suggested, arbitrarily chosen values (see caveat) for associated soil moisture transport properties.

index
soil texture
n
psi_s
K_s
b
comments
1
coarse0.421.03631.41E-54.26Loamy sand values*
2
medium/coarse0.434.14135.23E-64.74Sandy loam values*
3
medium0.439.35483.38E-65.25Loam values*
4
fine/medium0.404.13494.45E-66.77Sandy clay loam values*
5
fine0.465.26302.45E-68.17Clay loam values*
6
ice----------
7
organic0.439.35483.38E-65.25Loam values*
0
(ocean)----------

where

n is the porosity (dimensionless),
psi_s is the matric potential at saturation (in m)
K_s is the saturated hydraulic conductivity (in m/s), and
b (dimensionless) is the slope of the retention curve on a logarithmic graph, used to compute transport properties of subsaturated soils.
* CAUTION: The assignment of loamy sand transport parameter values to coarse soils does NOT imply that the "coarse" designation implies a loamy sand in the USDA soil texture triangle. Similarly, a "medium/coarse" designation does not imply a sandy loam, a "medium" designation does not imply a loam, and so on. The mapping of transport parameter values to soil texture in the table is highly arbitrary and technically incorrect. It is provided solely as a suggestion for the typical large scale (GCM) modeler, who could easily run into trouble if the "technically correct" numbers were used.

The suggested reclassification in the table reflects the inappropriateness of assigning hydraulic properties of soils as measured in the laboratory to GCM soil columns that represent extensive areas -- they tend to produce unrealistic resistance to soil moisture diffusion. This is almost certainly due to the inadequacy of current land surface models, which have very limited treatments of subgrid soil moisture variability, and to the fact that properties measured in the laboratory often do not describe soil behavior in the field, which is strongly influenced by spatial variability in texture, the presence of decayed root systems, wormholes, etc. As a makeshift response to this problem, a given soil type in the table above is arbitrarily assigned transport parameter values for a coarser textured soil. Determining the optimal parameter values for each type, which are probably very different from those listed above, would require much further research.

The values for the four transport parameters were obtained from the study of Cosby et al. (1984), who analyzed an extensive and diverse collection of soil samples.

C) SOIL PROFILE DEPTH. The soil profile thickness file was derived by Webb et al. (1991, 1993) from information contained in Volumes 2-10 of the FAO/UNESCO Soil Map of the World. First, the Earth was divided into nine continental regions: North America, Mexico/Central America, South America, Europe, Africa, South-Central Asia, North Central Asia, Southeast Asia, and Australia/South Asia. For each of these regions, the FAO records were examined to determine the profile thickness for a representative sample of every component soil type. When a thickness was undefined for a soil type, an arbitrary thickness of 3.6 meters was assigned; presumably the bedrock is at a greater depth than this. All soil elements of a given type within a given continental region were then assumed to have the same profile thickness. The thicknesses stored in the file's 1 degree x 1 degree array are the thicknesses for the dominant soil types within the grid squares, as determined by Zobler (1986).

D) AVERAGE SLOPE. The average topographical slope for each 1 degree x 1 degree square was derived from data sets constructed by the Science and Applications Branch of the EROS Data Center in Sioux Falls, South Dakota. Unlike the soil texture and soil profile thickness data, the average slope data reflects all of the soil regimes in a square, not just the dominant one. The slope estimates are crude, however, given the qualitative nature of the original data.

The Files

Format

Data Files

Name and Directory Information Naming Convention

The file naming convention for the FAO soil data set is

fao_soil.pppppp.1nnegl.ddd
where:
fao_soil = data product designator
pppppp = parameter name
types = soil type
textur = soil texture
depth = soil depth
slope = average slope
1 = number of levels
n = vertical coordinate, n= not applicable
n = temporal period, n = not applicable
e = horizontal grid resolution, e = 1 x 1 degree
gl = spatial coverage, gl = global (land)
ddd = file type designation, (bin=binary, ctl=GrADS control file)

Directory Path

/data/hydrology/soil/

Companion Software
Several software packages have been made available on the CIDC CD-ROM set. The Grid Analysis and Display System (GrADS) is an interactive desktop tool that is currently in use worldwide for the analysis and display of earth science data. GrADS meta-data files (.ctl) have been supplied for each of the data sets. A GrADS gui interface has been created for use with the CIDC data. See the GrADS document for information on how to use the gui interface.

Decompression software for PC and Macintosh platforms have been supplied for datasets which are compressed on the CIDC CD-ROM set. For additional information on the decompression software see the aareadme file in the directory:

software/decompression/

Sample programs in FORTRAN, C and IDL languages have also been made available to read these data. You may also acquire this software by accessing the software/read_cidc_sftwr directory on each of the CIDC CD-ROMs

The Science
Theoretical Basis of Data
The FAO Soil Map of the World is comprised of map units, which bound areas containing an association of soil units plus texture class and slope class. Each association of soil units can contain a dominant soil unit, associated soil unit and included soil unit. The area coverage, in a given map unit, for each of these categories are as follows:

Total area for each map unit is given, but area for each soil unit is not. The soil units are values representative of designated locations. The FAO has developed an algorithm for estimating the area of each soil unit within a soil association based on the number of soil units in each map unit.

The FAO system of soil classification has two levels, which are based on a inherent profile properties system. The highest level has 26 (used for soil type map) members and the lowest level has 106 members. Detailed profile descriptions, which are representative of the soil units, are given in the volume accompanying each map. The descriptions contain information on:

Textural classes reflect the relative proportions of clay (fraction less than 2 micrometers), silt (2-50 micrometers), and sand (50-2,000 micrometers) in the soil. The texture of a soil horizon is one of its most permanent characteristics. It is also a very important one because, in combination with other properties, it influences soil structure, consistence, porosity, cation exchange capacity, permeability and water holding capacity.

Three textural classes are recognized by the FAO Soil Map of the World:

  1. Coarse textured: sands, loamy sands, and sandy loams with less than 18 percent clay and more than 65 percent sand.

  2. Medium textured: sandy loams, loams, sandy clay loams, silt loams, silt, silty clay loams, and clay loams with less than 35 percent clay and less than 65 percent sand; the sand fraction may be as high as 82 percent if a minimum of 18 percent clay is present.

  3. Fine textured: clays, silty clays, sandy clays, clay loams, and silty clay loams with more than 35 percent clay.

The textural class refers to the texture of the upper 30 centimeters of the soil, which is important for tillage and water retention. The maps often state that a dominant soil type is composed of combinations of these textural classes (e.g., coarse AND medium for a given soil).

                                       100/\
                                         /  \
                                      90/    \10
                                       /      \
                                    80/        \20
                                     /          \
                     / \          70/            \30
                      |            /              \
                      |         60/                \40
                      |          /       FINE       \
                Percent clay  50/                    \50  Percent silt
                               /                      \         |
                            40/                        \60      |
                             /--------------------------\       |
                          30/                            \70   \ /
                           /                              \
                        20/                                \80
                         /--------          MEDIUM          \
                      10/         \                          \10
                       /    COARSE \                          \
                       ----------------------------------------\
                     100  90  80  70  60  50  40  30  20  10   
                                   <--------------
                                     Percent sand

To obtain the soil moisture transport parameters listed in the soil texture table, points corresponding to these textures or texture combinations were located on the U.S. Dept. of Agriculture (1951, p. 209) textural triangle, a rough reproduction of which is shown below:

                                       100/\
                                         /  \
                                      90/    \10
                                       /      \
                                    80/        \20
                                     /          \
                     / \          70/            \30
                      |            /              \
                      |         60/       C        \40
                      |          /                 /\
                Percent clay  50/\                /  \50  Percent silt
                               /  \              / SiC\         |
                            40/ SC \____________/______\60      |
                             /______\     CL    \ SiCL  \       |
                          30/ SCL    \___________\_______\70   \ /
                           /_________/          /         \
                        20/_         \    L    /   SiL     \80
                         /  \_   SL   \       /             \
                      10/\_   \_       \_____/         ______\90
                       / S \ LS \_          /         /  Si   \
                      /_____\_____\________/_________/_________\
                     100  90  80  70  60  50  40  30  20  10   
                                   <--------------
                                     Percent sand

The soil textures identified in the figure are:

C: Clay
SC: Sandy clay
SiC: Silty clay
SCL: Sandy clay loam
CL: Clay loam
SiCL: Silty clay loam
S: Sand
LS: Loamy sand
SL: Sandy loam
L: Loam
SiL: Silt loam
Si: Silt
The points were then arbitrarily shifted toward coarser soils, and transport parameters for the coarser soils were taken from Cosby et al. (1984), who used the same triangle to differentiate soil types.

Processing Sequence and Algorithms

Soil Type & Texture:

Zobler (1986) used a transparent overlay to subdivide the original source map into one-degree cells. The original source map (digitized from the 1:5,000,000 scale FAO Soil Map of the World) had a two degree grid size for latitudes below 60 degrees, a four degree grid size for latitudes between 60 and 80 degrees latitude, and a eight degree grid size above 80 degrees latitude. The soil type map was created by selecting the dominant soil unit, from the largest map unit covering the one-degree grid cell. Soil texture was derived from the soil profile information associated with each soil unit, using the methodology described in the SOURCE and THEORETICAL BASIS sections of this readme.

Soil Profile Depth:

Webb et al. (1991) developed the following set of decision rules for the soil depth data to standardize the data set, to check the data for errors and to correct them, and to fill in missing data:

Webb et al. (1991) reported that a number of soil types from each continent were completely missing depth data. To fill in these areas they substituted data using the same soil type from a different continental division. When data for a soil type were absent from all the continents, data were substituted from an adjacent soil type with similar descriptive characteristics.

Average Slope:

Arc/Info software was used for most processing steps in the construction of the slope data files generated by the EROS Data Center (which were then used to construct the average slopes), including projection from the bipolar oblique conformal projection to geographic (latitude-longitude) coordinates for the Americas. The remainder of the world was projected from the Miller oblated stereographic projection using software provided by Sprinsky (1992).

The average slopes were generated by some simple processing of the data sets produced by the EROS Data Center. These data sets provide, for each 1 degree x 1 degree square, the fractions f1, f2, and f3 of area covered by "level to gently undulating" (0-8%), "rolling to hilly" (8-30%) and "steeply dissected to mountainous" (>30%) slopes, respectively. For the calculation of the average slope, the 0-8% slope category was assigned a typical slope of 4%, the 8-30% slope category was assigned an average slope of 19%, and the >30% slope category was assigned the arbitrary slope of 40%. The average slope was then taken to be:

                     f1*4% + f2*19% + f3*40%
    average slope =  -------------------------
                     f1+f2+f3

Special Corrections/Adjustments:

A few of the 1 degree x 1 degree squares that were designated by Zobler (1986) as ice in the soil texture and soil type files are listed as land squares in the ISLSCP Initiative I vegetation data sets. To correct this inconsistency the "ice" squares, in the original soil data sets, that are designated as "tundra" in the vegetation data set have been changed to coarse (soil texture) and regosol (soil type).

The soil texture, depth, and slope data (that were originally archived on the ISLSCP Initiative I CD set) had the ISLSCP land/sea mask, applied to them. This mask had areas of land which were classified as ocean in the original Zobler (1986) Webb et al. (1991, 1993), and EROS data center files. The soil data on the ISLSCP Initiative I data was modified so that these grid cells were re-classified as land with an appropriate parameter value.

The soil data describe in this readme, and available as part of the Interdiscipline Data Collection (IDC) has been revised. Areas which were re-classified on the ISLSCP Initiative I soil data, as land, have been changed back to ocean on the IDC soil data. This process does not apply to the soil type data which never had the ISLSCP Initiative I land/sea mask applied to it.

The Goddard DAAC converted the soil files to IEEE floating point notation and re-oriented the soil type file to it's present form.

Scientific Potential of Data
The four soil data files are provided mainly for use in defining land surface properties for general circulation model (GCM) applications. Many land surface models coupled to GCMs require estimates of soil profile depth, surface slope, and soil moisture transport properties (as obtained from soil texture) for their runoff, soil moisture storage, and drainage parameterizations. Inherent in the data are large-scale spatial variations in the soil properties, which presumably are realistic even if values at various grid squares are inaccurate. This large-scale structure can be important for defining GCM climate.

Given that climate modelers are the expected users of the data, the danger of using the data for other applications must be stressed. Extracting a soil texture, slope, soil profile depth or soil type from the files for a specific small-scale region (even a region composed of numerous 1 degree x 1 degree squares) is foolhardy without further research into the reliability of the data in the region, as determined, for example, from the original FAO Soil Map of the World. At some squares, the data is undoubtedly unreliable. Even if the reliability were high, soil texture and profile depth are provided only for the dominant soil component of the 1 degree x 1 degree square, and thus the appropriate values in a subgrid region of interest can easily be missed. The moisture transport parameter values listed in the soil texture table are undoubtedly inaccurate and are provided ONLY to give climate modelers a consistent basis for performing intercomparison studies.

The data can be spatially aggregated by averaging the values in adjacent grid cells to create, for example, a 2x3 degree grid or a 3x5 degree grid. Although the grid cells are not equal area, and large errors would be introduced if a cell at the equator were averaged with a cell at the north pole, the errors from averaging adjacent cells will be within the accuracy limits for the data set.

Validation of Data
The original FAO data represent a generalization of more detailed data, which may be available in various countries, and which are in turn a generalized representation of reality. As stated by Zobler (1986), "about 11,000 maps were reviewed [to construct the FAO Soil Map of the World]; they varied widely in reliability, detail, precision, scales, methodologies, etc." As with any soil map, some of the variability in the actual soils is not shown on the map. Errors may have been introduced in the digitizing and map projection process.

The soil type, soil texture, and profile depth files contain data for the dominant soil type in each 1 degree x 1 degree square and thus ignore contributions from potentially significant secondary components. Soil texture, profile depth and average slope values were not available for all of the soil types mapped by Zobler (1986). The profile depths are generally based on depths measured for an equivalent soil elsewhere on the continent; depths are not actually measured at each square. As stated by Webb et al.(1991), the soil profile thickness in many cases represent minimum possible values because profile descriptions do not always extend to subsurface bedrock. The soil thicknesses range from 10 cm for Lithosol to 800 cm for Distric Nitosol in Africa. The spatial distribution of soil profile thickness can be summarized as thickest in the well-developed soils of tropical low latitudes and thinnest in the poorly developed soils of high latitudes. The soil profiles are thin in mountainous regions such as the Himalayas and Andes and are thick in mid-latitude peatlands such as those found in northern Europe and North America. For further discussion of the limitations of these data sets, see Zobler (1986) and Webb et al. (1991, 1993).

An obvious source of error in the average slope file is the arbitrary choice of 40% to represent all steep slopes, when all that is known is that they exceed 30%. Also, for the files used to compute the average slopes, assumptions were made on the percentage composition of the components. The vector data sets were gridded as separate data sets, and the data sets were merged in grid form. Some overlaps between data sets were removed manually.

Confidence Level/Accuracy Judgment:

Some measure of reliability was provided for the original FAO source maps, but these measures were not considered when constructing the soil texture, depth, and slope files, and corresponding arrays of reliability estimates are not available. The accuracy of the data is, of course, severely limited by the errors.

Measurement Error for Parameters and Variables:

The published FAO Soil Map of the World contains inset maps showing three categories of reliability for the source data used to make the map. Those interested in the reliability at a specific site should consult this source; again, digitized global reliability estimates are not available. Detailed soil surveys were performed only over selected areas of each continent.

Known Problems with the Data:

The FAO Soil Map of the World is becoming out-of-date because of recent soil surveys and new techniques for measurement and data handling. An international effort to develop a replacement, the Soil and Terrain (SOTER) digital data base of the world, is under development by the International Society of Soil Science, the International Soil Reference and Information Center, the FAO, and the UNEP.

Contacts

Points of Contact
For information about or assistance in using any DAAC data, contact

EOS Distributed Active Archive Center (DAAC)
Code 902.2
NASA Goddard Space Flight Center
Greenbelt, Maryland 20771
Internet: daacuso@daac.gsfc.nasa.gov
301-614-5224 (voice)
301-614-5268 (fax)

References
Cosby, B.J., G.M. Hornberger, R.B. Clapp, and T.R. Ginn, 1984. A statistical exploration of the relationships of soil moisture characteristics to the physical properties of soils, Water Resources Research, 20:682-690.

Food and Agriculture Organization (FAO) of the United Nations, 1970-78, Soil map of the world, scale 1:5,000,000, volumes I- X: United Nations Educational, Scientific, and Cultural Organization, Paris.

Sprinsky, William H., 1992. The inverse solution for the Miller oblated stereographic projection: Presented at the 27th International Geographical Congress, Washington, D.C.

U.S. Dept. of Agriculture, 1951. Soil Survey Manual. U.S. Dept. of Agriculture Agricultural Handbook, 18, 503pp.

Webb, R.S., C.E. Rosenzweig, and E.R. Levine, 1991. A global data set of soil particle size properties, NASA Tech. Memo. 4286, NASA, 34pp.

Webb, R.S., C.E. Rosenzweig, and E.R. Levine, 1993. Specifying land surface characteristics in general circulation models: soil profile data set and derived water-holding capacities, Global Biogeochemical Cycles, 7:97-108.

Zobler, L., 1986. A world soil file for global climate modeling. NASA Tech. Memo. 87802, NASA, 33pp.

Zobler, Leonard, 1987. A world soil hydrology file for global climate modeling: International Geographic Information Systems Symposium: The Research Agenda, November 15-18, 1987, Arlington, Virginia, Proceedings. 1:229-244.

NASA GSFC Goddard DAAC cidc site
NASAGoddardGDAACCIDC

Last update:Thu Jun 19 18:42:11 EDT 1997
Page Author: James McManus -- mcmanus@daac.gsfc.nasa.gov
Web Curator: Daniel Ziskin -- ziskin@daac.gsfc.nasa.gov
NASA official: Paul Chan, DAAC Manager -- chan@daac.gsfc.nasa.gov