Abstract In this work, we present an inverse computational method for the identification of the location, start time, duration and quantity of emitted substance of an unknown air pollution source of finite time duration in an urban environment. We considered a problem of transient pollutant dispersion under stationary meteorological fields, which is a reasonable assumption for the assimilation of available concentration measurements within 1 h from the start of an incident. We optimized the calculation of the source-receptor function by developing a method which requires integrating as many backward adjoint equations as the available measurement stations. This resulted in high numerical efficiency of the method. The source parameters are computed by maximizing the correlation function of the simulated and observed concentrations. The method has been integrated into the CFD code ADREA-HF and it has been tested successfully by performing a series of source inversion runs using the data of 200 individual realizations of puff releases, previously generated in a wind tunnel experiment.

中文翻译：

---

城市环境中点源未知有限持续时间空气污染物释放的逆识别

摘要 在这项工作中，我们提出了一种逆计算方法，用于识别城市环境中有限持续时间的未知空气污染源的位置、开始时间、持续时间和排放物质的数量。我们考虑了静止气象场下的瞬态污染物扩散问题，这是在事件开始后 1 小时内同化可用浓度测量值的合理假设。我们通过开发一种方法来优化源-受体函数的计算，该方法需要对与可用测量站一样多的后向伴随方程进行积分。这导致该方法的高数值效率。源参数是通过最大化模拟和观察到的浓度的相关函数来计算的。