Abstract A method is outlined and tested to detect low level nuclear or chemical sources from time series of concentration measurements. The method uses a mesoscale atmospheric model to simulate the concentration signature from a known or suspected source at a receptor which is then regressed successively against segments of the measurement series to create time series of metrics that measure the goodness of fit between the signatures and the measurement segments. The method was applied to radioxenon data from the Comprehensive Test Ban Treaty (CTBT) collection site in Ussuriysk, Russia (RN58) after the Democratic People's Republic of Korea (North Korea) underground nuclear test on February 12, 2013 near Punggye. The metrics were found to be a good screening tool to locate data segments with a strong likelihood of origin from Punggye, especially when multiplied together to a determine the joint probability. Metrics from RN58 were also used to find the probability that activity measured in February and April of 2013 originated from the Feb 12 test. A detailed analysis of an RN58 data segment from April 3/4, 2013 was also carried out for a grid of source locations around Punggye and identified Punggye as the most likely point of origin. Thus, the results support the strong possibility that radioxenon was emitted from the test site at various times in April and was detected intermittently at RN58, depending on the wind direction. The method does not locate unsuspected sources, but instead, evaluates the probability of a source at a specified location. However, it can be extended to include a set of suspected sources. Extension of the method to higher resolution data sets, arbitrary sampling, and time-varying sources is discussed along with a path to evaluate uncertainty in the calculated probabilities.

中文翻译：

---

通过表面浓度测量检测核试验：2013 年 2 月朝鲜地下试验放射性氙的分析

摘要 概述并测试了一种从浓度测量的时间序列中检测低水平核或化学源的方法。该方法使用中尺度大气模型来模拟来自已知或可疑来源的接收器的浓度特征，然后根据测量序列的片段连续回归，以创建时间序列的度量，用于测量特征和测量之间的拟合优度段。在朝鲜民主主义人民共和国（朝鲜）于 2013 年 2 月 12 日在丰溪附近进行地下核试验后，该方法被应用于俄罗斯乌苏里斯克（RN58）全面禁止核试验条约（CTBT）收集站的放射性氙数据。发现这些指标是一种很好的筛选工具，可以定位很有可能来自丰溪的数据段，特别是当相乘以确定联合概率时。来自 RN58 的指标也被用于计算 2013 年 2 月和 4 月测量的活动源自 2 月 12 日测试的概率。还对 2013 年 4 月 3 日/4 日的 RN58 数据段进行了详细分析，分析了丰溪周围的源位置网格，并将丰溪确定为最有可能的起源点。因此，结果支持强烈的可能性，即放射性氙在 4 月的不同时间从试验场发射出来，并在 RN58 处间歇性地检测到，这取决于风向。该方法不会定位未怀疑的源，而是评估指定位置处源的概率。但是，它可以扩展到包括一组可疑来源。将方法扩展到更高分辨率的数据集，