Data were analyzed to develop a model for predicting the incidence of immature rice kernels with white portions of the Hinohikari cultivar. The coefficient of determination (r2) between the incidence of white-based rice kernels and mean air temperature for 20 days after heading was high. Therefore, fitting a curve between these two parameters is a viable means of developing a prediction model for the incidence of white-based rice kernels. Milky white rice kernels are caused by high temperatures that accelerate the respiration rate of the rice plant, a lack of solar radiation, over-dense rice grains per m2 and low leaf blade nitrogen content, as well as heat that induces some ripening damage. Using these determinants, the following evaluation indicator was proposed:
DMG = SR × α × RUE/Gr
where SR is solar radiation, α the absorptance of solar radiation by the community of rice plants, RUE the radiation use efficiency and Gr the number of rice grains per m2. RUE could be influenced by air temperature and leaf blade nitrogen content. The relationship between air temperature and RUE was investigated considering the leaf color, which represents the leaf blade nitrogen content of the Koshihikari cultivar. RUE decreased with increasing air temperature and a high leaf color kept RUE high. The equations of the relationship between RUE and air temperature, and the leaf color value, express the influence of high temperature and leaf color on dry matter production. DMG of Hinohikari was successfully calculated by the same equation with an additional parameter. DMGHN for Hinohikari was plotted for the incidence of milky white rice kernels. The coefficient of determination between milky white rice kernels and DMGHN was high (r2 = 0.65). The following model used to predict the incidence of milky white rice kernels was obtained:
IMWK = 52.3 exp(-599.9 × DMGHN)
where IMWK was the incidence of milky white rice kernels (%).