The United Graduate School of Agricultural Sciences, Ehime University, 3-5-7 Tarumi, Matsuyama City, 790-8566, Japan.
Water Sci Technol. 2012;66(5):918-26. doi: 10.2166/wst.2012.256.
Suspended sediment, which is an important water quality characteristic concerning effluents from agricultural areas, was studied in relatively small rivers that drain agricultural watersheds with considerable rice paddy areas. Suspended sediment load (SL) was observed daily for thirty three months and analysed--applying data stratification. Suspended sediment prediction models were established and the effect of rice transplanting activities on the rivers' SL was estimated. Results showed that data stratification improved the discharge-SL correlation and reduced regression and curve-fitting errors, thereby improving the efficiency of the derived model equations. Clustering the months into the rice- and non-rice transplanting seasons also improved the resulting regression equations, although not statistically significantly. Suspended SL was found to be higher during the rice transplanting season and the activities contributed a considerable amount of suspended sediment during the period, supporting the conjecture that sediments come from sources other than natural soil erosion.
悬浮泥沙是与农业区排放物有关的重要水质特征，本研究针对的是那些相对较小、且排水流域内有大量稻田的河流。本研究对三十三个月的悬浮泥沙负荷（SL）进行了逐日观测，并采用数据分层分析的方法对其进行了分析。建立了悬浮泥沙预测模型，并估计了水稻移栽活动对河流 SL 的影响。结果表明，数据分层提高了流量-悬浮泥沙相关性，降低了回归和曲线拟合误差，从而提高了所得模型方程的效率。将月份聚类为水稻移栽季节和非水稻移栽季节，也改善了回归方程的结果，但没有统计学意义。研究发现，悬浮泥沙在水稻移栽季节较高，且该季节的活动带来了大量的悬浮泥沙，这支持了这样一种推测，即泥沙来自于自然土壤侵蚀以外的其他来源。
