How To Use the Model

 

Requesting a copy of DSSAT

The DSSAT shell was originally developed for a Windows environment, but can also be run in a Windows emulator on Linux and iOS platforms. If you do not have a copy of the DSSAT software, first request the software by selecting the “Download DSSAT” button and then completing the information requested. This is a two-step process; the system first checks to see if you have a valid e-mail address. Once your e-mail had been validated you are requested to complete a form. Upon review of your request you will receive a link to download the software. Unzip the file and install the DSSAT Software by selecting Install DSSAT. Note that you will need to install DSSAT with administrative privileges.

 

Agronomic Studies

DSSAT provides at least one real experimental data for each of the crops that can be simulated. This includes the daily weather data, soil surface and profile information, detailed crop management and some observations. More details can be found in the Minimum Data section. In order to run DSSAT for the example agronomy studies, open the DSSAT Shell from your desktop icon or App. In the middle panel you find the Selector for Crops. Select the type of crop you are interested in and then the specific crop. In the panel on the right-hand side a list of one or more experiments will appear. These are all real experiments that have been conducted in the past. Select the experiment you would like to simulate. By default all treatments associated with that experiment are selected. Then Click on the Run option on top of the DSSAT panel next to the lightning bold. A new window will open that will show all treatments that will be simulated. Select Run Model and the Cropping System Model of DSSAT will run in the background, starting at planting or the start of simulation date and ending when the model predicts harvest maturity or when the user has defined final harvest. The simulation of one single growing season should not take more than 1 second. Upon completion of all the simulations a “Simulations are completed” window pops up. Select the Analysis Tab and a list of output files will appear. For new users it might be best to select the “PlantGro.out” file first. This file includes daily outputs for most of the simulated plant growth variables, such as leaf, stem, seed, and root biomass, root length density for the different soil layers, leaf area index, drought stress and many others. Selecting “View” will show the detailed simulated output data, but “Plot” allows to display all data in a graphical format using the GBuild tool of DSSAT. Bolded variables mean that observed data from the experiment are available for a comparison between simulated and observed data. Select one or more variables you would like to plot as well as the treatments (“Runs”), and then click on “Next.” Both simulated and observed variables are shown, with the simulated data represented by the continuous line and observed by the symbols. Selecting “Statistic” will provide a statistical summary between simulated and observed. For time series data we prefer to use the “d-Statistic” and Root Means Square Error (RMSE).

DSSAT includes many different agronomic studies for most crops that have been conducted at many sites across the globe. For example, the maize model includes an Irrigation * Fertilizer experiment conducted in Gainesville, Florida, USA (UFGA8201), an Irrigation * Hybrid experiment conducted in Piracicaba, SP. Brazil (BRPI0202), a Temperature * CO2 experiment conducted in Griffin, Georgia, USA (GAGR0201), a Nitrogen * Phosphorus Experiment conducted in Wa, Ghana (GHWA0401), a Nitrogen * Hybrid experiment conducted in Waipio, Hawaii, USA ( IBWA8301), a Nitrogen * Plant Population experiment conducted in Ames, Iowa (IUAF9901MZ), and an irrigation experiment conducted for two years in Zaragosa, Spain (SIAZ9501 & SIAZ9601). Similar examples can be found for all other crops that are currently included in DSSAT. If you are interested in providing your data for inclusion in DSSAT, please do not hesitate to contact us. We are always interested in expanding the range of agronomic applications for different environments in order to help improve the performance of DSSAT and the Cropping System Model.

This Agronomic Studies section of DSSAT represented under the “Crops Selector” panel should only be used for crop model calibration, especially for the cultivar, variety, hybrid, or clone genetic coefficients, and model evaluation in response to different management inputs or environment. Any application should be conducted using the Seasonal, Sequence and Spatial Analysis options explained below.

 

Seasonal and Risk Analysis

Risk analysis is one of the main applications of DSSAT and allows users to evaluate alternate management practices for single growing seasons that account for both weather and economic uncertainty. The crop models were developed to address the Genotype * Environment * Management interactions. Using the seasonal analysis option of DSSAT, a user can compare the interaction of genotype and management for different environments, especially long-term historical weather data. In a standard risk analysis approach a user defines at least two or more management scenarios. Normally for the weather inputs at least 30 years of historical weather data are selected, analogous to climate normals that are based on 30 years of historical weather data. If long-term historical weather data are not available, a weather generator can be used. The Cropping System Model of DSSAT currently includes the WGEN and SIMMETEO weather generators. Externally generated weather data from MarkSim and other weather generators can also be used. The simulations are conducted for each unique combination of crop management and weather year. This provides a simulated distribution for yield, yield components, and other simulated variables. The economic uncertainty can be defined through prices files.

 

Rotation and Long-term Simulations

Cropping systems are not really defined by single growing seasons, but the long-term management practices that are implemented. This requires long-term simulations, starting with initialization of the cropping system environment with respect to soil water, nitrogen, phosphorus, potassium for the individual soil layers or horizon, soil surface residue and soil organic matter.

 

Spatial Analysis

DSSAT currently does not include tools that allow for the preparation of spatial input data and thematic map display of output data. Users should explore other tools, such as the CRAFT tool for spatial yield forecasting, MINK developed by the International Food Policy Research Institute (IFPRI), pDSSAT developed by the University of Chicago, and a new spatial modeling tool called DSSAT Pythia developed at the University of Florida.