Appendix VIII

The use of the soil data in CGMS

Introduction

Information associated with soil mapping units, or simply soil data, is used in two ways. Firstly it is used to determine the soil characteristics used for crop growth simulation such as: soil moisture content, seepage parameters, rooting depth, etc. Secondly, the soil data is used to determine whether or not to simulate growth for a particular crop. Ideally this decision would be taken on the basis of actual land-use information, but since this information is not generally available, a heuristic method is applied to determine the need for simulation. The decision to simulate is simply based on the suitability of a particular soil for a particular crop. If at least part of the soil mapping unit is deemed suitable then the simulation will be performed. GGMS yield figures are based on the suitable soils only!!.

A Soil mapping unit (SMU) is decribed by one or more sub-units (i.e. soil types). Depending on the applied map scale, these sub-units may not be mapped as individual cartographic units. These sub-units are called ‘soil typologic units’ (STUs) and describe the physical soil parameters. This information is stored in the SOIL_ASSOCIATION_TABLE table, which contains the following fields:

SMU_NO:	SMU number
STU_NO:	STU number
PERCENTAGE:	percentage area of SMU covered by this STU

 

System wide soil parameters

In the current CGMS version, some information that is closely linked to the soil model is not differentiated with respect to the soil types, but is considered to be ‘system wide’. This information is stored in table SITE, which, contains the following fields:

DD:	depth of drainage (cm) (not used)
IDRAIN :	presence of drains (not used)
IFUNRN:	used in calculation of preliminary infiltration rate
NOTINF:	used in calculation of preliminary infiltration rate
ZTI:	initial depth of groundwater table (cm)
WAV:	moisture content above wilting point (cm)
IZT :	influence of groundwater (0 == no, 1 == yes)
MAX_SURFACE_STORAGE:	maximum surface storage capacity (cm)

It should be noted that the current system wide settings of IFRUNF = 0 and NOTINF = 0 result in a calculation of the preliminary infiltration rate equal to the rainfall plus the surface storage. However, the setting of MAX_SURFACE_STORAGE = 0 means that the surface storage is always 0. Also IZT = 0 means that no groundwater is considered, and consequently setting ZTI has no effect.
 

Simulation parameters and soil characteristics

Description of soil characteristics for the crop simulation model in CGMS only relies on two parameters. These two parameters, the so-called rooting depth (RD) and soil physical group (SPG), describe the soil for simulation purposes. The rooting depth parameter refers to a so-called rooting depth class. These classes are described by a minimum depth. The depth values for each class are stored in a table called ROOTING_DEPTH, and describe the depth to which the soil allows root growth. The table contains the following fields:

CLASS:	class number
MIN_DEPTH:	minimum rooting depth in cm.

The soil physical group parameter refers to one of a number of possible sets of soil parameters. These parameters are stored in a table called SOIL_PHYSICAL_GROUP, and the values in this table describe the soil characteristics as used CGMS. The operational CGMS version at the JRC only contains 7 different soil physical groups. The table consists of:

SOIL_GROUP_NO:	the soil physical group identifier	[-]
HYDR_CONDUCT_SATUR:	hydraulic conductivity saturated soil	[cm d-1]
MAX_PERCOL_ROOT_ZONE:	maximum percolation rate root zone	[cm d-1]
MAX_PERCOL_SUBSOIL:	maximum percolation rate subsoil	[cm d-1]
RQ_MOISTURE_DEFICIT_DEEP:	required moisture deficit deep seedbed	[-]
SEEPAGE_1_SHALLOW:	1st topsoil seepage parameter for shallow seedbed	[-]
SEEPAGE_2_SHALLOW:	2nd topsoil seepage parameter for shallow seedbed	[-]
SEEPAGE_1_DEEP:	1st topsoil seepage parameter for deep seedbed	[-]
SEEPAGE_2_DEEP:	2nd topsoil seepage parameter for deep seedbed	[-]
CRITICAL_AIR_CONTENT:	critical soil air content for aeration	[cm3cm3]
SOIL_MOISTURE_CONTENT_SAT:	soil moisture content at saturation	[cm3cm3]
SOIL_MOISTURE_CONTENT_WP:	soil moisture content at wilting point	[cm3cm3]
SOIL_MOISTURE_CONTENT_FC:	soil moisture content at field capacity	[cm3cm3]

Each soil typologic unit can now be described by a combination of these two values RD and SPG, and in fact the minimum required columns in the SOIL_TYPOLOGIC_UNIT table are:

STU_NO:	identifier for the soil typologic unit
CALCULATED_ROOTING_DEPTH:	rooting depth CLASS
SOIL_GROUP_NO:	soil group number

Rooting depth values and the soil physical group have to be derived from information associated with the soil map. This derivation is soil map specific and depends on the applied legend. It should be noted that the number of soil physical groups is not fixed or limited to 7. If soil characteristics cannot be described by any of the available soil physical groups, a new soil physical group can be constructed. It should also be noted that CGMS as a result of the applied method to describe crop transpiration is more sensitive to rooting depth variation than to variation in soil physical group.
 

Suitability and simulation units

As mentioned above, the system will simulate crop growth for those crops and soil combinations that are deemed suitable. This suitability is not expressed in terms of the actual crops, but in terms of a so-called crop-group. Every crop is labelled as belonging to one of a number of crop groups. Currently the system recognises two crop groups, root crops (e.g. potato and sugar beet) and cereal crops (e.g. wheat, barley, and maize). The SUITABILITY table contains every valid combination of STU number and crop group number. The columns of the SUITABILITY table are:

CROPGROUP_NO:	normally 1 or 2
STU_NO:	identifier for the soil typologic unit

Suitability of a STU_NO for a CROP_GROUP is determined by many factors that are derived from the soil type. In general it can be said that it depends on alkalinity, salinity, rooting depth and soil texture. Note that suitability may also depend on local conditions and soil types that would not normally be considered suitable for a crop-group could be considered suitable depending on local usage. This is especially an important consideration for regional applications. Another point to note is that SPG/RD combinations derived for the STU are partly independent from the suitability derived from the STU. In other words, two different STUs can yield the same SPG/RD combination but one STU may be considered suitable whereas the other may not be suitable for a crop-group.

Crop growth simulations are carried out for the elementary mapping units. Because meteorological input data is assumed to be the same for all EMUs in a grid, the result of the simulation for a grid depends only on the SPG/RD combinations for which the simulation is carried out. Since many STUs have the same SPG/RD combination, a table is introduced that contains the unique and suitable SPG/RD combinations for every grid. This table, SIMULATION_UNIT is used only to reduce program execution time and does not reflect any model characteristics. The columns of the simulation_unit table are:

CROP_NO:	the crop identifier
GRID_NO:	identifier for the grid
SOIL_GROUP_NO:	the soil group number
CALCULATED_ROOTING_DEPTH:	the rooting depth CLASS

 

Aggregation

The model output, which is calculated for a combination of soil physical group and rooting depth as defined by the simulation units, needs to be ‘aggregated’ in order to describe the simulated results at EMU level. Each EMU by definition is covered by one unique SMU. Each SMU consists of one or more STUs and the relative contribution of every STU to the SMU in terms of area is described in the SOIL_ASSOCIATION_COMPOSITIONtable. It is assumed that the same distribution of STUs holds for the EMU. As stated before, the defining factors for the STU with respect to the simulation are the rooting depth (RD) and the soil physical group (SPG). The simulation results for the EMUs are derived from the simulation unit (SPG/RD) results by weighing these results with the relative contribution of the STUs that corresponds to this SPG/RD combination.

A naïve approach would be to just multiply a STU result with its percentage and sum the results for all STUs within the SMU. Looked at in another way, calculate the potential production for each STU, sum the results and divide by the EMU area to get the potential EMU yield.

Unfortunately, this would not give the desired result. To understand this, we have to look a few steps ahead, and look at the level at which we could conceivably calculate a production. This we can only do when we know the actual acreage used for a particular crop, e.g. obtained from agricultural statistics. This acreage however is not just an area. It is a ‘qualified’ area. It is qualified by the mere fact that the crop has actually been cultivated there and therefore it is a ‘suitable’ area. If we were to calculate the potential EMU yield in the naïve way, then this potential yield would also take into account the non-suitable areas. The result would be that the production for an EMU, or the production for an area consisting of several EMUs would be underestimated. It is therefore necessary to ‘qualify’ the area in the calculation of the potential EMU yield by dividing the ‘naïve’ result by the fraction of suitable area within the EMU. E.g. if only 80% of the EMU area contributes then we divide the ‘naïve’ result by 0.8 to find the ‘qualified’ potential yield. In order to increase processing speed, the suitable area for a given crop-group for every SMU is stored in the SMU_SUITABILITY table.