Study on monitoring of cultivation and manure in resources circulation type farm.

近年来，河流水质在畜牧业地区已成为一个问题，需要掌握实际状况。本文调查了青森县畜牧业集水区流出的水对水质的影响。结果表明，总氮浓度与草地面积比成正比，并随外流速率的增加而增加。草地总氮含量为3.3%。结果表明，融雪过程中SS、TP和DTP的负荷均超过90%。此外，研究还表明，SS、TP和DTP主要在白天排放，而DTN主要在夜间排放。六个重要的草化学成分估计使用传感系统与高光谱成像传感器安装在拖拉机上。采用多元线性回归方法建立校正模型，并采用3种统计分析方法评价准确度。这些方法是交叉验证的标准误差（SECV），决定系数和交叉验证使用的偏倚。采用评价指数（EI）对校正模型进行评价。结果表明，传感系统估计草的化学成分具有良好的潜力。

Recently, the river water quality has becomes a problem in the livestock farming areas and the actual state grasping is demanded.In this paper, the effect of outflow water from catchment areas in livestock farming on water quality ware investigated in Aomori Prefecture. As a result, total nitrogen concentration was proportional to the grassland area ratio and increased and outflow rate of. total nitrogen from the grassland was 3.3 %. It has been shown that the load of SS, TP and DTP exceed 90 percent in snowmelt. In addition, it has been shown that SS, TP and DTP were discharged during the daytime and DTN was discharged during the nighttime.A system was developed to estimate grass chemical composition values at the field level. Six important grass chemical compositions were estimated using a sensing system with a hyper-spectral imaging sensor mounted on a tractor. Calibration models were obtained by using multiple linear regression, and accuracy was evaluated with three statistical analysis methods. These methods were the standard error of cross validation (SECV), the coefficient of determination, and the Bias used by cross-validation. The calibration models were evaluated the evaluation index (El). Results indicated good potential for the sensing system to estimate grass chemical composition.