Replication package for "Regional trade and sustainable intensification advance rice self-sufficiency in East Africa"

Optimization model for sustainable intensification of agriculture under regional trade This replication package contains the code and data required to reproduce the analysis presented in the article "Regional trade and sustainable intensification advance rice self-sufficiency in East Africa". The model maxmizes farmers' benefits subject to resource constraints. Scripted routines and optimization model The main model is written in GAMS. R scripts are used for pre- and post-processing data. Current versions of GAMS and R are required for replication. Run the contained scripts in the following sequence: Use MAIN_data_processing.R to pre-process input data, then convert it to GAMS data format by running input_xlsx_to_gdx.gms Run the baseline optimization in BASE_rice_model.gms with fixed scenario to determine water use in this fixed setting. Use water_constraint.R to scale the water use from baseline and use it in the actual opitmization model. Run the main optimization model in rice_model.gms. Due to the significant amount of scenarios this can take some time. Use gdx_to_csv.gms to convert data from GAMS to CSV fomrat to make it readable with other programmes, including R. Use reproduce_figures_numbers.R to analyze optimization model outputs and reproduce numbers, percentages and figures included in the article. This package provides the following datasets: Input datasets Custom Global Agro-Ecological Zones version extended for double-cropping and high spatial resolution of rice in GAEZ_rice_data.zip Productions costs assesmbled from literature and processed for use in this modeling framework in production_costs.csv subbasin_to_marketshed.csv and subbasin_to_country.csv provide mapping of spatial units for use in the GAMS model. These datasets need to be downloaded to fully reproduce the analysis: Supplementary data from Traberend et al. (2021) at: https://doi.org/10.1016/j.oneear.2021.02.012 Producer Prices from FAOSTAT Rice areas from SPAM: https://doi.org/10.7910/DVN/FSSKBW The economic model in GAMS can be run without the full data integration by using the processed data file provided in: model_inputs_cleaned.xlsx A sensitivity check using a variation of income elasticities to compute demand can be run by instead using the processed input data in model_inputs_POS_INC.xlsx Output data Output data of the economic optimization is provided in the csv files in results_csv.zip. This archive contains the main results used in the article and the outputs for model runs for the sensitivity checks using alternative income elasticities reported in the Supplementary.区域贸易下农业可持续集约化优化模型 本复现包包含复现《区域贸易与可持续集约化提升东非水稻自给率》一文分析所需的代码与数据。 该模型以农户收益最大化为目标，同时受资源约束条件限制。 脚本程序与优化模型 核心优化模型采用GAMS语言编写，数据预处理与后处理环节使用R脚本完成。复现该分析需使用当前最新版本的GAMS与R软件，请按照以下顺序运行附带脚本： 1. 运行`MAIN_data_processing.R`完成输入数据预处理，并通过执行`input_xlsx_to_gdx.gms`将数据转换为GAMS支持的GDX格式。 2. 执行`BASE_rice_model.gms`运行基准情景优化，确定固定情景下的水资源使用量。 3. 运行`water_constraint.R`基于基准情景的水资源使用量进行缩放，并将其应用于实际优化模型。 4. 执行`rice_model.gms`运行主优化模型。由于情景数量较多，该步骤可能耗时较长。 5. 运行`gdx_to_csv.gms`将GAMS输出的数据转换为CSV格式，以便其他程序（包括R）读取。 6. 运行`reproduce_figures_numbers.R`分析优化模型输出结果，复现论文中的数值、占比与图表。 本复现包提供以下数据集： ### 输入数据集 1. `GAEZ_rice_data.zip`：经扩展适配双季稻种植、具备高空间分辨率的定制化全球农业生态区（Global Agro-Ecological Zones, GAEZ）水稻专用数据集。 2. `production_costs.csv`：从公开文献汇总整理，并经本建模框架适配处理的生产成本数据集。 3. `subbasin_to_marketshed.csv`与`subbasin_to_country.csv`：为GAMS模型提供子流域与销售区、子流域与国家的空间映射关系数据集。 需另行下载方可完整复现分析的数据集： 1. Traberend等人（2021）的补充数据：https://doi.org/10.1016/j.oneear.2021.02.012 2. FAOSTAT发布的生产者价格数据 3. SPAM提供的水稻种植面积数据：https://doi.org/10.7910/DVN/FSSKBW 若无需完整数据集成流程，可直接使用`model_inputs_cleaned.xlsx`中提供的预处理数据文件运行GAMS经济模型。 若需通过调整收入弹性参数变体计算需求以开展敏感性分析，可改用`model_inputs_POS_INC.xlsx`中的预处理输入数据。 ### 输出数据集 经济优化模型的输出数据已打包至`results_csv.zip`压缩包中，该存档包含论文使用的核心结果，以及正文补充材料中报告的、采用替代收入弹性开展的敏感性检查模型运行输出。