The article develops a new method of spatiotemporal modelling of possible locations of ragweed areas and estimation of pollen emission intensity of these areas. The method uses spline approximation with automatic computations of typical pollen production models in the R language. It applies a number of approaches eliminating the uncertainty of various parameters that should be taken into consideration for this modelling. The authors have consolidated some experimental data on the amount of Ambrosia pollen in the air depending on environmental conditions, and found some new possible ragweed pollen sources on the territory of the city of Vinnytsia and the adjacent villages. The study used spline approximation of the first (lines) and second (parabolas) orders were applied. Data of bi-hourly pollen concentrations were obtained at the pollen monitoring site of National Pirogov Memorial Medical University (Vinnytsia). To perform the approximation, wind speed, wind direction and relative humidity were taken into the consideration. The authors used the polar coordinate system to formalize the problem. PRA Finder computer program was developed, which has been successfully tested on real data and for which a certificate of copyright registration has been obtained in Ukraine. As a result of the method application, the known areas were corroborated and new possible plots were revealed. The comparison of the forecasted areas for 2014 with the actual ones showed a close match. However, it should be taken into consideration that the proposed method is viable only when applied under the condition that the wind velocity does not exceed 4 m/sec and the relative humidity is between 20 and 90%.

本文开发了一种新方法，对豚草区域的可能位置进行时空建模并估算这些区域的花粉排放强度。该方法使用样条近似与 R 语言中典型花粉生产模型的自动计算。它应用了多种方法来消除建模时应考虑的各种参数的不确定性。作者整合了一些关于Ambrosia数量的实验数据空气中的花粉取决于环境条件，并在文尼察市和邻近村庄的领土上发现了一些新的可能的豚草花粉来源。该研究使用了第一（线）和第二（抛物线）阶的样条近似。在国家皮罗戈夫纪念医科大学（文尼察）的花粉监测点获得每两小时花粉浓度的数据。为了进行近似，考虑了风速、风向和相对湿度。作者使用极坐标系将问题形式化。开发了 PRA Finder 计算机程序，该程序已在真实数据上成功测试，并已在乌克兰获得版权注册证书。作为方法应用的结果，已知区域得到证实，新的可能地块也被揭示出来。2014 年的预测面积与实际面积的比较显示出高度匹配。但需要注意的是，该方法仅适用于风速不超过 4 m/sec 且相对湿度在 20% 至 90% 之间的条件下才可行。