空间插值法在人群包虫病患病率预测中的应用

目的探索不同空间插值方法在人群包虫病患病率预测中的应用并比较其效果。方法以四川、西藏、甘肃和青海四省(区)青藏高原地区2012年全国包虫病流行情况调查获得的124个藏区县的人群包虫病患病率数据为基础,使用Arcgis10.0软件,采用局部多项式插值法、反距离权重插值法和kriging插值法进行空间插值,比较青海省治多县和玛多县人群包虫病患病率推算的效果。结果 3种方法预测的精度各有优势,对青海省治多县和玛多县人群包虫病患病率误差最小的为Kriging插值方法,预测的患病率分别为2.04%和5.28%。结论空间插值方法可用于青海省治多县和玛多县人群包虫病患病率预测,以Kriging插值法的预测效果较好。

Use of spatial interpolation to predict the prevalence of hydatidosis

Objectives To explore the use of different methods of spatial interpolation to predict the prevalence of hydatidosis and to compare their efficiency. Methods Data on the prevalence of hydatidosis in 124 Tibetan counties in Sichuan, Tibet, Gansu, and Qinghai provinces/autonomous regions were collected from the National Sampling Survey on the Prevalence of Hydatidosis in 2012. Arcgis10.0 software was used to compare the predicted rates of hydatidosis in Zhidoi County and Maduo County in Qinghai Province according to spatial interpolation. Three methods of spatial interpo lation were used： local polynomial interpolation, inverse distance weighting, and kriging. Results The three methods have an advantage in terms of predictive accuracy. Results indicated that local polynomial interpolation had the largest predictive error while kriging had the smallest. Thus, kriging was chosen. According to kriging, the predicted rates of hydatidosis in Zhidoi County and Maduo County were 2. 04% and 5.28%, respectively. Conclusion Spatial interpolation can be used to effectively predict the rates of hydatidosis in Qinghai Province. Kriging was more effective at prediction than two other methods of spatial interpolation.