Development of mpi_EPIC model for global agroecosystem modeling

Kang, Suhyun; Wang, D.; Nichols, Jeffrey A.; Schuchart, Joseph; Kline, Keith L.; Wei, Yaxing; Ricciuto, D. M.; Wullschleger, Stan D.; Post, Wilfred M.; Izaurralde, R. C.

doi:10.1016/j.compag.2014.12.004

Cited by 8 publications

(7 citation statements)

References 20 publications

(19 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…HPC clusters). Specifically, MPI has been used successfully in other applications for parallelizing crop models (Nichols et al, 2011;Kang et al, 2015;Jang et al, 2019;Vital et al, 2013). The message-passing parallel programming model allows multiple independent DSSAT-CSM instances to run concurrently, each of which manages its own memory and most of its input and output.…”

Section: Parallelizationmentioning

confidence: 99%

Parallel gridded simulation framework for DSSAT-CSM (version 4.7.5.21) using MPI and NetCDF

Alderman¹

2021

Preprint

View full text Add to dashboard Cite

Abstract. The Decision Support System for Agrotechnology Transfer Cropping Systems Model (DSSAT-CSM) is a widely used crop modeling system that has been integrated into large-scale modeling frameworks. Existing frameworks generate spatially-explicit simulated outputs at grid points through an inefficient process of translation from binary, spatially-referenced inputs to point-specific text input files followed by translation and aggregation back from point-specific, text output files to binary, spatially-referenced outputs. The main objective of this paper was to document the design and implementation of a parallel gridded simulation framework for DSSAT-CSM. A secondary objective was to provide preliminary analysis of execution time and scaling of the new parallel gridded framework. The parallel gridded framework includes improved code for model-internal data transfer, gridded input/output with the Network Common Data Form (NetCDF) library, and parallelization of simulations using the Message Passing Interface (MPI). Validation simulations with the DSSAT-CSM-CROPSIM-CERES-Wheat model revealed subtle discrepancies in simulated yield due to the rounding of soil parameters in the input routines of the standard DSSAT-CSM. Utilizing NetCDF for direct input/output produced a 3.7- to 4-fold reduction in execution time compared to text-based input/output. Parallelization improved execution time for both versions with between 12.2- (standard version) and 13.4-fold (parallel gridded version) speedup when comparing 1 to 16 compute cores. Estimates of parallelization of computation ranged between 99.2 (standard version) and 97.3 percent (parallel gridded version) indicating potential for scaling to higher numbers of compute cores.

show abstract

Section: Parallelizationmentioning

confidence: 99%

Parallel gridded simulation framework for DSSAT-CSM (version 4.7.5.21) using MPI and NetCDF

Alderman¹

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The MPI specification was selected primarily due to its widespread usage and suitability for distributed memory systems (e.g., HPC clusters). Specifically, MPI has been used successfully in other applications for parallelizing crop models (Nichols et al, 2011;Kang et al, 2015;Jang et al, 2019;Vital et al, 2013). The message-passing parallel programming model allows multiple independent DSSAT-CSM instances to run concurrently, each of which manages its own memory and most of its input and output.…”

Section: Parallelizationmentioning

confidence: 99%

“…As noted by Kang et al (2015) and Jang et al (2019), examples of improving execution time for agricultural system models being applied at large scales using HPC are limited. Nevertheless, considerable increases in speed of execution have been documented in some models by leveraging parallel execution and linking to gridded input-output libraries.…”

mentioning

confidence: 99%

“…For example, Nichols et al (2011) found a 40-fold speedup of simulations by porting the Erosion Productivity Impact Calculator (EPIC) model from Windows to Linux and parallelizing simulations (i.e., running multiple instances of a model simultaneously) on an HPC cluster. Subsequent improvements using the Message Passing Interface (MPI) have permitted up to 512-fold speed-up of EPIC simulations (using 2048 parallel processes) with nearly linear scaling up to 128 parallel processes (Kang et al, 2015). Zhao et al (2013) reported an average 20-fold speed-up on multi-core Windows computers while running the Agricultural Production Systems Simulator (APSIM) model using the HT Condor grid computing middleware.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Parallel gridded simulation framework for DSSAT-CSM (version 4.7.5.21) using MPI and NetCDF

Alderman

2021

Geosci. Model Dev.

View full text Add to dashboard Cite

Abstract. The Decision Support System for Agrotechnology Transfer Cropping Systems Model (DSSAT-CSM) is a widely used crop modeling system that has been integrated into large-scale modeling frameworks. Existing frameworks generate spatially explicit simulated outputs at grid points through an inefficient process of translation from binary spatially referenced inputs to point-specific text input files, followed by translation and aggregation back from point-specific text output files to binary spatially referenced outputs. The main objective of this paper was to document the design and implementation of a parallel gridded simulation framework for DSSAT-CSM. A secondary objective was to provide preliminary analysis of execution time and scaling of the new parallel gridded framework. The parallel gridded framework includes improved code for model-internal data transfer, gridded input–output with the Network Common Data Form (NetCDF) library, and parallelization of simulations using the Message Passing Interface (MPI). Validation simulations with the DSSAT-CSM-CROPSIM-CERES-Wheat model revealed subtle discrepancies in simulated yield due to the rounding of soil parameters in the input routines of the standard DSSAT-CSM. Utilizing NetCDF for direct input–output produced a 3.7- to 4-fold reduction in execution time compared to R- and text-based input–output. Parallelization improved execution time for both versions with between 12.2- (standard version) and 13.4-fold (parallel gridded version) speed-up when comparing 1 to 16 compute cores. Estimates of parallelization of computation ranged between 99.2 % (standard version) and 97.3 % (parallel gridded version), indicating potential for scaling to higher numbers of compute cores.

show abstract

“…The use of renewable energy sources reduces greenhouse gas emissions (Garlucci et al, 2015;Harrous et al, 2017). Agricultural production is a complex agroecosystem (Belcher et al, 2004;Kang et al, 2015;Golub et al, 2017), which is constantly affected by human production activities (Conway, 1987;Moonen and Barberi, 2008;Preston et al, 2015). Traditionally, the structure of the agroecosystem is made up of crop production (based on crop rotation) and livestock (Harrous et al, 2017;Golub et al, 2020a).…”

Section: Introductionmentioning

confidence: 99%

Researching of indicators of agroecosystem without external energy supply

Golub¹,

Chuba²,

Lutak³

et al. 2021

j. cent. eur. agric.

View full text Add to dashboard Cite

The agroecosystem without external energy supply can provide compensation for the needs of the agroecosystem in liquid fuel, heat and electricity at the expense of plant biomass as a renewable energy source. Based on the analysis of the energy balance of the agroecosystem, the area under rapeseed needed for biodiesel production and the area under crops needed for bioethanol production were calculated for the needs of the agroecosystem, as well as additional quantities of straw to compensate for the loss of humus in crop rotation, which sets the amount of plant biomass that can be used to produce heat and electricity using cogeneration gas plants. It was found that the average typical needs of the agroecosystem in diesel fuel from 60 to 80 kg/ha, and the ratio of the area under rapeseed needed for biodiesel production for the needs of agroecosystem to the total area of crop rotation with grain yield of canola from 1500 to 2500 kg/ha is from 8.3 to 18.5%. At the average typical demand of the agroecosystem for gasoline is from 15 to 16 kg/ha, the ratio of the area under wheat, barley and corn required to produce bioethanol for the needs of the agroecosystem to the total area of crop rotation with a crop yield of 3000 to 5000 kg/ha is from 0.7 to 0.9%. The increase in grain yield from 3000 to 5000 kg/ha causes an increase in the supply of organic matter to soil to a greater extent than its mineralization, and accordingly, the deficit of organic matter and humus decreases from 507 to 180 kg/ha, the volume of plant biomass (mainly straw) to compensate for the deficit of organic matter and humus decreases from 1867 to 217 kg/ha, which causes an increase in the possible volumes of plant biomass for energy production for sale from 27 to 2963 kg/ha.Respectively, electricity production increases from 48 to 2301 kWh/ha and heat power from 0 to 25,5 GJ/ha, while providing its own needs for heat and electricity by burning biogas in a cogeneration plant.

show abstract

Development of mpi_EPIC model for global agroecosystem modeling

Cited by 8 publications

References 20 publications

Parallel gridded simulation framework for DSSAT-CSM (version 4.7.5.21) using MPI and NetCDF

Parallel gridded simulation framework for DSSAT-CSM (version 4.7.5.21) using MPI and NetCDF

Parallel gridded simulation framework for DSSAT-CSM (version 4.7.5.21) using MPI and NetCDF

Researching of indicators of agroecosystem without external energy supply

Contact Info

Product

Resources

About