pycgms 0.1.0

The PyCGMS package provides a python implementation of the crop simulation

pyCGMS

The European system for crop monitoring and yield forecasting

The Joint Research Centre of the European Commission runs an operational system for crop monitoring and yield forecasting called MARS (Monitoring Agricultural ReSources). MARS is used within agricultural monitoring activities applied to Europe, sub-Saharan Africa and other areas of the world. Crop yield forecasting is undertaken to provide monthly bulletins forecasting crop yields to support the EU's Common Agriculture Policy (CAP). Providing early warning of crop shortages or failure provides rapid information for EU development aid activities to support food insecure countries.

The MARS system implements different components for monitoring including meteorological data, crop simulation modelling, satellite data and field vists. Finally, statistical post-processing is used in combination with regional yield statistics to make forecasts of actual crop yield and production.

The Crop Growth Monitoring System (CGMS)

The crop simulation component in MARS is implemented by a system called CGMS (Crop Growth Monitoring System) which includes the WOFOST crop simulation model and was originally implemented in C++. CGMS links to a database structure which provides the inputs (weather data, crop parameters, soil types and crop calendars) and stores the output (simulated biomass, yield, leaf area index, etc) from the crop simulation model. Several version of this CGMS database structure exist: V8, V12 and the current V14.

The C++ implementation of CGMS is proprietary, owned by EC-JRC and is not available for download. In recent implementations of CGMS, the C++ version of C++ is superseded by the crop simulation models implemented in the BioMA framework.

A python version of CGMS

The PyCGMS package provides a python implementation of the crop simulation system embedded in the Crop Growth Monitoring System (CGMS). Under the hood, the actual crop simulations are carried out by the WOFOST implementation in PCSE which provides a fully open source implementation of many crop simulation models developed in Wageningen.

PyCGMS was designed to be compatible with all versions of the CGMS database and can therefore also run on legacy CGMS implementations. The original C++ CGMS executable also provided functionality for interpolating weather data. pyCGMS does not provide this functionality and it will not do so in the future as dedicated packages are available for processing of weather data.

Limitations

PyCGMS is very much a work in progress and therefore does not yet provide all functionality that was implemented by the C++ CGMS version. Notably, PyCGMS does not provide support for writing output to the CGMS database structure. This is also a design choice because CGMS provides large amount of simulation output that can be better loaded into the database using dedicated loaders. However, such loaders are not yet available.

Installation and usage

First of all, PyCGMS requires a [CGMS database] structure in order to retrieve its input data. A definition of the structure can be found in JRC wiki using the link above. An example CGMS8 database in SQLite can be found here.

Next, a python environment must be created for PyCGMS. The environment for PyCGMS is similar to the environment required for PCSE, so please have a look at the detailed instructions there. After installing PCSE into the environment PyCGMS can be installed with:

pip install pycgms

Note that depending on the database system you are using, additional database drivers may need to be installed.

Using PyCGMS can be started using the commandline script pycgms. Use:

$ pycgms --help
usage: pycgms [-h] --db_version {8,12,14} --dsn DSN --crop CROP --year YEAR
              [--grid GRID] [--run_till yyyy-mm-dd]
              [--aggr_level {stu,smu,grid}] --output OUT_PATH
              [--output_type {csv,xls,hdf5,json}]
              [--use_isw_date USE_ISW_DATE]

Run a gridded WOFOST simulation on a CGMS database.

optional arguments:
  -h, --help            show this help message and exit
  --db_version {8,12,14}
                        Type of CGMS DB to use (either 8, 12 or 14).
  --dsn DSN             SQLAlchemy connection URL for CGMS DB to connect to.
                        See also
                        http://docs.sqlalchemy.org/en/latest/core/engines.html
  --crop CROP           Run simulations for given crop number.
  --year YEAR           Run simulations for given year. The year refers to the
                        year in the crop_calendar table which usually
                        indicates the year where sowing of emergence takes
                        place.
  --grid GRID           Run simulations for given grid. Optional, by default
                        all grids will be simulated where the crop is defined.
  --run_till yyyy-mm-dd
                        Run simulations up till this date. This is useful for
                        simulations in the current year where not all weather
                        data are available up till the end of the simulation.
  --aggr_level {stu,smu,grid}
                        Aggregation level for output, default is "stu"
  --output OUT_PATH     Store simulation results at this location.
  --output_type {csv,xls,hdf5,json}
                        Type of output file to write
  --use_isw_date USE_ISW_DATE
                        If True the start_date from the table
                        INITIAL_SOIL_WATER will be used ascampaign_start_date,
                        default False.