Development of an orchestration aid system for gridded crop growth simulations using Kubernetes

Highlights

• An orchestration aid system (OAS) was developed to support a gridded crop growth simulations.

• The OAS requires minimum labor and expertise in management and operation of cluster computers.

• Functionality of the OAS was demonstrated using single board computers.

• The OAS would help researchers design site-specific options for climate change adaptation.

Abstract

Spatial simulations of crop growth under climate change have been limited to researchers who have access to the resources for high performance computing. The objective of this study was to develop an orchestration aid system for concurrent gridded simulations of crop growth, which would support the design of climate change adaptation options on crop production without expertise in distributed computing. The orchestration aid system was designed to help a user build a set of virtualized cluster computers using a simple input file, which would require little expertise in distributed computing, rather than manual configuration. The orchestration aid system, which was referred to as GROWLERS-kube, was implemented to launch multiple sets of gridded simulations using pods or containers managed by Kubernetes. As a case study, GROWLER-kube was executed using 16 Raspberry Pi 4 computers to perform 120 sets of the gridded simulations under diverse crop management options, including varying planting date and cultivar, for the period from 2001 to 2010. The wall time or the elapsed time for the given sets of the gridded simulation differed by configuration of virtualized cluster computers, such as the number of pods used for server and client nodes, although the total number of physical nodes were identical. For example, the wall time difference between virtualized cluster computer sets was about 28.9% when 15 worker nodes were used. In particular, the acceleration of the gridded simulations was at maximum using a large number of the virtualized cluster computers with a small number of nodes. It was found that the spatial distribution of planting dates and cultivars was similar to that of a previous study based on field experiments mostly in regions where rice is usually grown. These results suggest that GROWLERS-kube would facilitate the spatial assessment of climate change impact on crop production without considerable effort and expertise in distributed computing, which would aid a researcher to focus on the design of adaptation strategies.

Introduction

Spatial assessment of climate change impact on crop production would aid identification of the adaptation measures suitable for a region of interest (Rosenzweig et al. 2014). It would be prudent to take into account both the outcome of local studies and crop growth simulations for the reasonable design of adaptation options (Beveridge et al. 2018). In particular, it would be preferable to use crop growth simulations over an extended area in order to overcome the limited number of field-based studies. For example, the gridded simulations of crop growth have been performed at global and regional scales (Asseng et al. 2019), which provides spatial information on the potential outcomes of adaptation options under future climate conditions.

A process-based crop model, which simulates the biophysical responses of crops to given conditions (He et al. 2017), has been used to evaluate the management practices chosen under a given climate condition. For example, Bassu et al. (2009) suggested that planting date shift would be a viable option to minimize the negative impact of climate change on crop production using crop growth simulations. Chun et al. (2016) illustrated that the benefit of planting date shift for rice production would differ by region through gridded crop growth simulations over Southeast Asia.

A large number of crop growth simulations would be needed to evaluate crop management options under environmental and management conditions (Jang et al. 2019). Stakeholders or local farmers would benefit from knowledge of planting dates and cultivars suitable for a given climate condition (Sacks et al. 2010). Still, the adaptation measures, e.g., planting date and cultivar, may differ by time and locality (Ishigooka et al. 2017). As a result, the number of crop growth simulations increases considerably to evaluate adaptation options over extended time periods and regions. This would require considerable computing resources and expertise for a large amount of the simulations. For example, Xiong et al. (2020) used a high performance computer to launch 1440 sets of the global gridded simulations with the combinations of climate projections, crop models, parameter strategies and sowing dates.

Various software has been developed to support the gridded simulations (Resop et al., 2016, Hyun et al., 2017, Yoo et al., 2018). Researchers can write their own scripts to deal with the tasks for the gridded simulations such as preparation of inputs, execution of model run and aggregation of outputs. Alternatively, integrated tools for the gridded simulations can be used to minimize efforts for such tasks (Shelia et al. 2019). However, these tools are often developed for a small set of gridded simulations, which supports limited functionality to minimize labor or accelerate computation for a large number of the gridded simulations.

Multiple sets of gridded crop growth simulations can be performed in parallel using the orchestration of workloads. The orchestration of the gridded simulations include configuration, coordination and management of containers, which enable customized environment and performance isolation (Rodriguez and Buyya 2019). For example, containers specialized for the gridded simulations can be deployed to a large number of computers using an orchestration system such as Kubernetes (Beltre et al. 2019). However, the demand for computer expertise could make it challenging for researchers to perform multiple sets of gridded crop growth simulations using the orchestration system.

The objective of our study was to develop a software framework that supports the orchestration of the gridded crop growth simulations. We also focused on implementation of the framework for researchers who have limited budget and expertise in the orchestration system. In particular, the orchestration aid system was applied to single board computers, which allowed for a considerably large number of crop growth simulations with minimum costs and efforts. Such a system would facilitate the spatial assessment of climate change impact on crop production and the design of adaptation strategies in a region.