The Fourth Assessment Report (AR4) of the Intergovernmental Panel on Climate Change (IPCC) projects that a rise in the global mean temperature due to anthropogenic greenhouse gas (GHG) emissions will have a large effect on crop yields. Crop yields change mainly due to the influence of agroclimatic factors, such as temperature, precipitation and insolation, and technological factors, such as agrichemicals, fertilizer use, cultivar improvement, and irrigation water supply. Crop yields have continued to increase with sustainable investment in and development of crop cultivation technologies since human beings began cultivating agricultural crops. However, some regions are projected to experience a large decrease in crop yields in the future due to climate change. Therefore, an understanding of the regions that are vulnerable to climate change through quantitative assessments of the projected changes in crop yields is necessary.

The impacts of climate change on crop yields have been assessed by impact simulations, in which climate projections, obtained from simulations using General Circulation Models (GCMs), are input into crop models, which estimate crop yields, using predominantly information about climatic and soil conditions. Unfortunately, the GCM climate projections developed by research institutes around the world vary in their results, due to differences in the semi-empirical parameters and the process formulations used for each model; therefore, there are uncertainties in the outputs of the impact assessment models. Quantifying the uncertainty of the estimates of changes in crop yields due to differences in the climate projections is of great importance to aid in the understanding of which regions are most vulnerable to future climate change and in using this understanding as supporting information to investigate effective measures against climate change.

Impact assessment studies on the changes in crop yields under climate change have been carried out at the regional or global levels based on the characteristics of the models used and the purposes of each study. Except for some comprehensive studies, only a small number of GCM climate predictions and socioeconomic scenarios were used in these impact assessment studies due to the large data size of the climate projections and the extended computation time. The impact of climate change on crop yields is closely related to global food security. Because the agricultural producing countries in the Asia-Pacific region are vulnerable even under the current climate system and extreme events, crop yield change is projected to have serious impacts under future climate change. A quantitative assessment is necessary to determine which climatic elements influence yield changes and to ascertain the amount of uncertainty associated with using multiple climate projections for a study region. Given this background, the purpose of this study is to quantitatively assess the change in rice yields using a global scale crop model, considering the uncertainty of climate change projections using data from the Coupled Model Intercomparison Project (CMIP5) GCM climate projections under 4 RCP scenarios (RCP 2.6, RCP 4.5, RCP 6.0 and RCP 8.5) for 3 time periods (2020s, 2050s and 2080s) for agricultural countries in the Asia-Pacific region.

The M-GAEZ model was modified for better assessment of climate change impacts from the GAEZ model, which is an model for estimating potential crop yields at a global scale that was developed by the Food and Agriculture Organization (FAO) and the International Institute for Applied Systems Analysis (IIASA). The potential crop yields are calculated based on conditions, such as the climate, soil, and input level. Specific climate conditions, such as the daily mean temperature [°C], daily precipitation [mm/day], daily mean irradiation [W/m²], and daily mean windspeed [m/s], are used as input information. The M-GAEZ model can estimate the potential yields for 8 varieties of rice. Some varieties of rice differ in their growth periods, depending on the cultivated area. Four varieties (growing periods of 105, 120, 135 and 150 days) for Japonica and four varieties (105, 120, 135 and 150 days) for Indica were treated distinctively. We then calculated the potential rice yields based on the rice variety at each site with the 1990s (1991-2000) as the reference years and determining the most productive variety and its planting day. To conduct the impact assessment, the M-GAEZ model requires daily climate scenarios at a spatial resolution of 1º×1º latitude/longitude for each climate condition as input data. We created each mesh of a daily climate scenario using the GCM climate projections. The crop yield calculation flow consists of 4 modules: the climate module, the crop growth module, the soil constraint module, and the CO2 fertilization module. The crop growth and soil constraint modules were based on formulating the existing M-GAEZ model. For the CO2 fertilization effect, multipliers intended specifically for each plant type were used to consider the fertilization effect. The results obtained from this study are summarized below.

After calculating each mesh of rice yield from multiple climate projections based on the simulation periods, we calculated the country-specific yield average using data on the area of rice cultivated areas and country-boundary data to find the rate of rice yield change by country. Changes in rice yields were predicted for major 20 countries located in the Asia-Pacific region. As a result, after we considered adaptation measures, in Russia and North Korea, which are located at relatively high–latitudes for the region, rice yields are predicted to increase in the 2020s, 2050s and 2080s, under all RCP scenarios, both in the current rice cultivated area and relative to the 1990s, under all emissions scenarios, due to temperature rise. It was predicted that rice yields will increase more as we approach the end of the 21st century, the RCP 8.5 scenario, in particular. When we did not consider adaptation measures, it was predicted that rice yields decreased in most countries, except for the districts in high latitude areas, by many GCMs. After having predicted rice yield changes for each GCM according to RCP scenarios, there was still a large of uncertainty in the predicted results, due to the choice of GCM climate projections. When we did not consider adaptation measures, the uncertainty ranges were 13.5% (-5.7% to -19.2%) in the 2020s, 18.0% (-4.2% to -22.2%) in the 2050s, and 37% (-7.1% to -44.1%) in the 2080s under the RCP 8.5 scenario. As mentioned above, because the prediction results for rice yield change vary greatly depending on the choice of the GCM, uncertainty of prediction must be considered when examining climate change adaptation policy in many countries. In addition, 14% (-10.8% to -24.9%) of the uncertainty range was predicted, due to the choice of an RCP scenario in the 2080s, when we did not consider adaptation measures.

1. 서론

2. 연구 자료 및 방법

2.1 기후데이터

2.2 글로벌 작물 수확량 예측 모델

3. 연구내용

3.1 기후변화에 따른 쌀 수확량 전망

3.2 쌀 수확량 예측 불확실성 평가

3.3 쌀 생산 적지 이동 분석

4. 결론 및 토론