Abstract: Objective To establish a route sampling method based on Monte Carlo simulation with the use of software programming, and to preliminarily analyze the effects of different route sampling methods on the mosquito monitoring results by the route index method.Methods The Python 3.80 software was used to create Monte Carlo simulations of the walking paths of monitoring personnel for route index mosquito monitoring in various residential areas. The simulations were performed with random route sampling and spatially stratified route sampling separately, and 1 000 different routes were simulated for each preset termination value. Excel 2019 software was used for data entry and arrangement. The route index, absolute error and variance of the simulation monitoring results were calculated, and the number of buildings sampled in the study area was counted. SPSS 26.0 software was used for statistical analysis. The data used for Monte Carlo simulations were the standing water data obtained from a survey of Aedes albopictus breeding sites in a large residential area (402 000 m²) in Shanghai by visual inspection in May and July 2023.Results Based on the survey results of mosquito breeding sites in May and July 2023 in the study area, Monte Carlo simulation was carried out with the termination presets of 50, 100, 150, 200, 250, and 300 households within the buffer zone. The simulation results show that the sampling probability of each building in the study area is not the same when sampling with random path, and the sampling times of the middle buildings are more. With the increase of sample size, the absolute error and variance decrease gradually. The comparison between the two sampling methods showed that when the sample size was the same, the efficiency of spatially stratified sampling was all greater than 1.25, indicating spatially stratified sampling showed higher efficiency than random route sampling.Conclusions A research method for route sampling was developed based on Monte Carlo simulation by using Python to simulate physical walking, and the optimizing walking paths can improve the sampling efficiency for the route index method. The route index method based on spatially stratified sampling can obtain the more representative Ae. albopictus monitoring results.

Key words: Route index, Aedes albopictus, Monte Carlo simulation, Spatial sampling

摘要： 目的 基于蒙特卡罗模拟建立使用软件编程的路径抽样研究方法,初步分析不同路径抽样对路径指数法蚊虫监测结果的影响。方法 使用Python 3.80软件编写代码,基于蒙特卡罗模拟监测人员行走方式,该方法可适用于各种居民区开展路径指数蚊虫监测。模拟采用随机路径抽样和空间分层路径抽样2种方式,对每种终止预设值分别模拟1 000次不同的路径。采用Excel 2019软件进行数据录入和整理,计算模拟蚊虫监测结果的路径指数、绝对误差和方差,统计研究区域中建筑被抽样次数,采用SPSS 26.0软件进行统计分析。蒙特卡罗模拟所使用的数据为2023年5和7月通过目测法对上海市某大型居住区(40.20万m²)开展白纹伊蚊孳生地调查所得的积水数据。结果 以研究区域2023年5和7月孳生地调查结果为总体,以缓冲区范围内有50、100、150、200、250和300户为终止预设值开展蒙特卡罗模拟。模拟结果显示,以随机路径抽样时,研究区域中每处建筑被抽样的概率并不相同,中部建筑被抽样次数较多;随着样本量的增加,绝对误差和方差逐步减小。2种抽样方法比较显示,当样本量相同时,空间分层抽样效率均>1.25,空间分层抽样效率高于随机路径抽样。结论 基于Python代码模拟人工行走,建立了采用蒙特卡罗模拟路径抽样的研究方法,发现优化路径行走方式可提高路径指数法的抽样效率,基于空间分层抽样的路径指数法,可获得更具代表性的白纹伊蚊监测结果。

关键词: 路径指数, 白纹伊蚊, 蒙特卡罗模拟, 空间抽样