In this study, we try to analyze the recipe structure composed dishes, focusing the eating phase which is an important phase of at-home food consumption. When the appearing frequency of dishes at a dinner is computed using a dish data base, SHOKU-MAP, monitors have beverage and rice at almost every dinner. We calculated the matrix of dish pairs, after specifying high appeared dishes and high appeared dish pairs per annum. It was possible to visualize the state of recipe appearance about 84 monitors taking over 6 high appeared dishes, by drawing the graph replacing the matrix. We observed that the graph was related with cooking motivation by using quantification theory type Ⅱ, categorizing the monitor to two types by numerical data specified by dividing the plot field of the graph to meshes. We confirmed that there was the relationship between the recipe graph and the recipe structure which was influenced by cooking motivation. It was suggested that there was the relationship between the household's birthplace and the recipe graph, by analyzing the relationship between the recipe graph and the family profile.
We developed a model for estimation of diurnal variation of water temperature in a paddy field with an emphasis on field-use, and we evaluated this model by comparing the data measured at 2 paddy fields managed by 2 different farmers. In addition, the factors influencing estimation error were examined and the feasibility of the model was discussed in terms of its estimation accuracy. The diurnal variation estimated by the developed model was in good agreement with the measured values. Further, the root mean squared error (RMSE) and maximum RMSE were around 1.2 °C and 2.26 to 2.58 °C, respectively. The results showed that the model has sufficient accuracy to provide useful information for water management in a paddy field. However, on several days, the values estimated at around 4:00am and around 3:00pm had large errors of 3 °C to 4 °C, respectively: these errors were due to the approximation of soil temperature by using a sinusoidal wave and the use of constant values for vegetation cover ratio, which neglected the effect of solar altitude. In addition, the estimated values were equal to the temperature of irrigation water if the depth of the flooding water was close to 0, because this model neglects the time delay until the level of irrigation water reaches the point for measuring water temperature. This resulted in a large error of almost 5 °C at a maximum between the measured and estimated values. However, in this study, the literature values were mainly used as the model parameters. Regarding the input values in the time series, the depth of flooding water and the flow rate of irrigation water were estimated on the basis of the water-management records obtained by farmers, and the soil temperature was also estimated. This approach reduced the data-collection items required for water-temperature estimation.
To conduct the technical examinations for estimating rice planted area using high resolution X band SAR satellite imagery data, we obtained COSMO-SkyMed imageries (SPOTLIGHT-2 mode) and carried out field survey at the flat area in Japan. As the results, following points were clarified. 1) Critical height of rice plant for specular reflection: Under the condition of X-HH at low incidence angle, back scatter was slightly observed on the imagery of 25 cm height (above water surface) of rice plant. 2) Practical usage of composite imagery for image interpretation under different incidence angles: The imageries under different incidence angle (23.7 and 56.1 degree) were composed to examine the misregistration. The composed image can be used for actual image interpretation, since the degree of misregistration is limited. Through the examinations, it was clarified that the important knowledge as a decision standard for acquisition period and conditions to estimate rice planted area using X band SAR satellite in the flat area.
The objective of this study was to obtain basic data, in order to discuss the type of farming in which rice is produced sustainably in rice terraces. Working hours, production costs, and income associated with rice production were surveyed on two farmers (farmer A, farmer B) in Hoshino Village, Fukuoka Prefecture, Japan. As a result, the working hours per 10 ares for farmers A and B were 3.2 times and 2.1 times longer, respectively, than the national average. Production costs, including interest expense and land rent, per 10 ares were 1.7 times (farmer A) and 1.5 times (farmer B) higher than the national average, and labor cost accounted for a large proportion of this cost. The income per 10 ares for both the farmers surveyed was significantly high2.2 times (farmer A) and 3.8 times (farmer B) higher than the national average. This is because the farmers sold their rice products at high prices. On the other hand, the income per day for farmer A was lower than the national average, and in the case of farmer B was only 1.7 times higher than the national average, due to the long working hours. The reproducible prices of brown rice, which are the sales prices required for each farmer to produce rice sustainably, were calculated as 557 yen/kg (farmer A) and 380 yen/kg (farmer B). Actual sales prices were 333 yen/kg (farmer A) and 425 yen/kg (farmer B), and farmer B earned a high income by selling rice products at a higher price than the reproducible price. The survey results indicated that it is necessary to reduce production costs and properly determine sales prices of rice products for sustainable rice production in rice terraces. Therefore, it is important for each of the farmers to reduce labor hours and intensify sales efforts by increasing the added value of rice.