Impacts of Climate Change on Japanese Sweet Potato Yield: Time Series Analysis of Process-Based-Model Building and Regional Differences

Yuki KIMURA

DOI: https://doi.org/10.2457/srs.52.293

2022-01-01

Studies in Regional Science

Abstract:Numerous papers evaluate the impacts of climate change on crop yield by applying two types of methods. One is a simulation analysis with process-based-models that represents the key processes governing how plants grow. The other is a statistical model estimation to obtain parameters that represent the yield of a plant with variables of interest such as temperature, precipitation, etc. Process-based-models cost more to build than statistical models. Therefore process-based-models are usually used for major crops. As for sweet potato, there are few process-based models. Statistical models have advantages in that they cost less than process-based-models and consider unknown processes that a process-based model cannot handle. Preceding studies showed there is an effect of CO2 on fertilization in many crops, such as rice, wheat and potato. However, whether CO2 increases the yield of sweet potato is not clear. Sweet potato production is concentrated in the Kanto and Kyushu areas of Japan. In Kanto, sweet potato is mainly used for raw consumption, whereas in Kyushu the main use of sweet potato is alcohol or starch production. There is also a distinction in potato variety used and climatic conditions between the Kanto and Kyushu areas. Therefore, the impacts of climate change on sweet potato yield likely differ between the Kanto and Kyushu areas. This study first investigated how sweet potatoes react to CO2 concentrations that is not well considered in existing process-based-models about sweet potato. Second, I considered regional differences in the climatic impacts on sweet potato production in the main sweet potato production areas of Japan, Kagoshima and Ibaraki Prefectures. The results of analysis suggest there is no clear cut relationship between yield of sweet potato and CO2 concentrations. This implies no supportive evidence to add CO2 to the input of a process-based model. Second, after confirming that CO2 does not have an effect on sweet potato yield, this study rebuilt statistical models without CO2. These models implied that Kagoshima Prefecture is more sensitive to a rise in temperature than Ibaraki Prefecture, and both prefectures have experienced decreasing yields after the end of the 20th Century, Kagoshima Prefecture appears to have a more severe downward change.JEL Classifications:Q54, C51