| Abstract: | Abstract: The need for solar radiation (Rs, MJ m–2 d–1) estimation remains a common concern for agronomists. Evaluation of crop productivity is primarily based on Rs data, which are difficult to collect because of cost and calibration requirements. Generally, historical Rs data are more difficult to obtain. This study focused on an estimation model based on the daily range of temperature and evaluated its accuracy from the viewpoint of crop productivity analysis. The variability of an empirical coefficient in the model (Krs), which was derived from the relation between Rs and daily range of temperature (Tmax – Tmin) was analyzed using climatic data observed in Japan considering data availability and quality. Krs had significant monthly differences, and it significantly increased from 1981 – 1985 to 2003 – 2007 at all 10 locations. Period-month interactions were not significant, except for in Utsunomiya, suggesting that the seasonal pattern did not change during the period. Weather data indicated that the increase in Krs was caused not only by increased solar radiation but also by a decrease in Tmax – Tmin. The substantial differences in Krs produced considerable bias for the estimated Rs when the estimation was conducted with a constant Krs (0.16). Despite the bias, the model is considered to perform well given the present availability of Rs data. The results of this study suggest that the evaluation of the seasonal pattern of Krs greatly improves the model accuracy. |
