Global Rice Land Suitability and Adaptation Strategies Under Climate Change

This study evaluated the land suitability of irrigated and rainfed rice cultivation under current and projected climate change scenarios for the year 2050 on a global scale. , the climate and soil factors utilized for this purpose were obtained from the WorldClim website (https://www.worldclim.org/) and represent the historical monthly weather data for the period between 2001 and 2021 (https://worldclim.org/data/monthlywth.html). Moreover, future climate projections were sourced from the same source for the period 2041 to 2060 (centered around 2050).The projections are based on mean value of ten Global Climate Models (GCMs) from the CMIP6 dataset and two scenarios (SSP245 and SSP585), including models such as ACCESS-CM2, CanESM5-CanOE, EC-Earth3-Veg, FIO-ESM-2-0, GISS-E2-1-H, HadGEM3-GC31-LL, INM-CM4-8, IPSL-CM6A-LR, MRI-ESM2-0, and UKESM1-0-LL (https://worldclim.org/data/cmip6/cmip6_clim2.5m.html). The future climate scenarios were developed using a new set of integrated assessment models (IAMs) that incorporate the Shared Socioeconomic Pathways (SSPs) and Representative Concentration Pathways (RCPs), as outlined by O'Neill et al. (2016). Two integrated scenarios were considered in this study. SSP245 (a combination of SSP2 with RCP4.5) and SSP585 (a combination of SSP5 with RCP8.5). SSP2 represents a scenario characterized by a continuation of existing social, economic and technological trends, with minimal deviation from historical patterns. In contrast, SSP5 depicts a pathway of fossil-fuelled development, characterized by rapid technological advancement and human capital growth. With regard to radiative forcing, RCP4.5 represents a medium scenario (4.5 W m⁻² by 2100), whereas RCP8.5 represents a high-emission scenario (8.5 W m⁻² by 2100). To ensure the highest possible spatial accuracy, 2.5 minutes (~21 km2 at the equator) spatial resolution was applied to all maps. Furthermore, maps of rice cultivation area was obtained from the MAP SPAM2020 (Spatial Production Allocation Model) website (https://www.mapspam.info), which provides detailed spatial data on global rice production.Soil parameters were then determined for each point using soil information obtained from the FAO soil map. The FAO soil map provided vital parameters such as organic carbon (OC), exchangeable sodium percentage (ESP), cation exchange capacity (CEC), CaCO3, pH, EC which extracted from the Harmonized World Soil Database (HWSD) at a scale of 1:5,000,000, sourced from the FAO Soil Portal website. (Supplementary Figure 2). The slope map was created using the digital elevation model (for more details, see Guerra et al., 2020). The SPAM2020 for rainfed and irrigated rice was used to intersect the climate zones and soil layers, focusing specifically on the areas dedicated to this area. 

The final land suitability map was generated by overlaying the suitability output with raster files representing both irrigated and rainfed harvested areas, thus creating a composite map of land suitability . This stage ensured the precise delineation of areas suitable for rice cultivation. Following the creation of land suitability maps for a variety of scenarios and adaptation strategies, the results were compared in order to identify the optimal strategy. The final map identifies the most efficacious adaptation strategy for enhancing land suitability in each region in the context of projected climate change. Also, The overview of the status of rainfed and irrigated rice worldwide.

本研究在全球尺度下，评估了当前及2050年气候变化预估情景下，灌溉稻与雨养稻的种植土地适宜性。本研究所用的气候与土壤因子取自WorldClim网站（https://www.worldclim.org/），其数据为2001-2021年的历史逐月气象数据（https://worldclim.org/data/monthlywth.html）。此外，2041-2060年（以2050年为核心时段）的未来气候预估数据同样取自该平台。该预估数据基于CMIP6数据集下10个全球气候模式（Global Climate Models, GCMs）的平均值，以及两种情景（SSP245与SSP585），涉及的模式包括ACCESS-CM2、CanESM5-CanOE、EC-Earth3-Veg、FIO-ESM-2-0、GISS-E2-1-H、HadGEM3-GC31-LL、INM-CM4-8、IPSL-CM6A-LR、MRI-ESM2-0及UKESM1-0-LL（https://worldclim.org/data/cmip6/cmip6_clim2.5m.html）。未来气候情景由一套全新的综合评估模型（integrated assessment models, IAMs）构建，该模型融合了共享社会经济路径（Shared Socioeconomic Pathways, SSPs）与典型浓度路径（Representative Concentration Pathways, RCPs），相关框架由O'Neill等人于2016年提出。本研究共考虑两种综合情景：SSP245（SSP2与RCP4.5的组合）与SSP585（SSP5与RCP8.5的组合）。其中SSP2情景延续当前社会、经济与技术发展趋势，与历史模式偏差极小；而SSP5则代表化石燃料驱动的发展路径，以技术快速进步与人力资本增长为核心特征。就辐射强迫而言，RCP4.5为中等排放情景（至2100年达4.5 W·m⁻²），而RCP8.5则为高排放情景（至2100年达8.5 W·m⁻²）。为保障最高空间精度，所有地图均采用2.5角分（赤道处约21 km²）的空间分辨率。此外，水稻种植面积数据取自MAP SPAM2020（空间产量分配模型，Spatial Production Allocation Model）网站（https://www.mapspam.info），该平台提供全球水稻生产的精细化空间数据。随后，本研究通过联合国粮食及农业组织（Food and Agriculture Organization, FAO）土壤图获取的土壤信息，为每个栅格点计算土壤参数。该FAO土壤图提取自1:500万比例尺的世界土壤协调数据库（Harmonized World Soil Database, HWSD），并提供有机碳（organic carbon, OC）、交换性钠百分比（exchangeable sodium percentage, ESP）、阳离子交换量（cation exchange capacity, CEC）、CaCO3、pH值、EC等关键土壤参数，数据取自FAO土壤门户网站（补充图2）。坡度图通过数字高程模型（digital elevation model）生成，详细信息参见Guerra等人2020年的研究。本研究将雨养稻与灌溉稻的SPAM2020数据与气候分区、土壤图层进行空间叠加分析，重点聚焦于水稻种植专属区域。

最终的土地适宜性地图通过将适宜性输出结果与代表灌溉稻与雨养稻收获面积的栅格文件叠加生成，从而得到土地适宜性综合地图。该步骤实现了水稻种植适宜区域的精准划定。在针对多种情景与适应策略生成土地适宜性地图后，本研究对结果进行对比以筛选最优适应策略。最终地图可识别出在预估气候变化背景下，各区域提升土地适宜性的最有效适应方案。此外，本研究还提供了全球雨养稻与灌溉稻的种植现状概览。