Spatial Database Management System (SDBMS) and attribute data sets. These data
sets have the smallest possible number of spatial units and attributes required for a
practical assessment and evaluation.
The SDBMS stores the minimum spatial datasets in a shape file format, including administrative boundaries (ATHAxx.shp), cropping areas (Cxx.shp) and rice
(as used in our paper), as well as a soil series map (SOILxx.shp) and a weather zone
map for the ECHAM4 climate model (WSTAxx.shp), where xx is the administrative code of a given level, i.e., country, province etc. In the implementation, the four
spatial data layers of a given administrative boundary are overlaid to create
Simulation Mapping Units (SMU), each with a unique administrative code, land
use code, soil series code and weather zone code. These spatial data layers must be
prepared and overlaid using Geographic Information System (GIS) software.
The core attribute databases of the CropDSS tool include the soil attribute data
sets (Vearasilp and Songsawat 1991), the genetic coefficients of rice varieties, and
the measured or generated or climate model scenario weather data grids. The rice
genetic coefficients data for this experiment came from DSSAT MDS, based on
field experiments conducted in Thailand, while weather data was obtained from
SEA START RC at Chulalongkorn University, also in Thailand. These data sets
had a simple text file format so new data could be entered directly into the system
(Hoogenboom et al. 2003).
10.3.1.3 CSM-DSSAT Model Coupling
CropDSS was developed under the loose coupling approach, the aim being to avoid
redundant programming. The individual CSM-DSSAT model was coupled at the
SMU level based on a vector file format and set forth in reference to uniform soil
series and weather zone maps for selected administrative boundaries (Sui and
Maggio 1999; Hartkamp et al. 1999).
By executing a batch file, CropDSS executes the CSM-DSSAT model for each
SMU, one by one. After one simulation, a set of “DSSAT output files” is generated,
and for an “output data translation module”, output variables, such as yield,
evapotranspiration rates and crop water productivity, are written into a “summary.out output file”. Each line of the “summary.out output file” presents output
variables for one simulation. This output file is then used to generate “GIS output
maps”, such as yield, water and nitrogen maps. These maps can be visualized in
CropDSS and saved in a shapefile format for future use.
386
C. Marohn et al.
sets have the smallest possible number of spatial units and attributes required for a
practical assessment and evaluation.
The SDBMS stores the minimum spatial datasets in a shape file format, including administrative boundaries (ATHAxx.shp), cropping areas (Cxx.shp) and rice
(as used in our paper), as well as a soil series map (SOILxx.shp) and a weather zone
map for the ECHAM4 climate model (WSTAxx.shp), where xx is the administrative code of a given level, i.e., country, province etc. In the implementation, the four
spatial data layers of a given administrative boundary are overlaid to create
Simulation Mapping Units (SMU), each with a unique administrative code, land
use code, soil series code and weather zone code. These spatial data layers must be
prepared and overlaid using Geographic Information System (GIS) software.
The core attribute databases of the CropDSS tool include the soil attribute data
sets (Vearasilp and Songsawat 1991), the genetic coefficients of rice varieties, and
the measured or generated or climate model scenario weather data grids. The rice
genetic coefficients data for this experiment came from DSSAT MDS, based on
field experiments conducted in Thailand, while weather data was obtained from
SEA START RC at Chulalongkorn University, also in Thailand. These data sets
had a simple text file format so new data could be entered directly into the system
(Hoogenboom et al. 2003).
10.3.1.3 CSM-DSSAT Model Coupling
CropDSS was developed under the loose coupling approach, the aim being to avoid
redundant programming. The individual CSM-DSSAT model was coupled at the
SMU level based on a vector file format and set forth in reference to uniform soil
series and weather zone maps for selected administrative boundaries (Sui and
Maggio 1999; Hartkamp et al. 1999).
By executing a batch file, CropDSS executes the CSM-DSSAT model for each
SMU, one by one. After one simulation, a set of “DSSAT output files” is generated,
and for an “output data translation module”, output variables, such as yield,
evapotranspiration rates and crop water productivity, are written into a “summary.out output file”. Each line of the “summary.out output file” presents output
variables for one simulation. This output file is then used to generate “GIS output
maps”, such as yield, water and nitrogen maps. These maps can be visualized in
CropDSS and saved in a shapefile format for future use.
386
C. Marohn et al.