In this work we study different characteristics of fog long-term persistence, in events with different physical formation mechanisms. Specifically, we focus on the characterization of fog long-term persistence from observational data, by means of a Detrended Fluctuation Analysis (DFA) of its associated low-visibility time series. We analyze fog events with radiation and orographic underlying physical formation mechanisms, and identify a two-range pattern of long-term persistence. Our analysis leads to the emergence of a characteristic time, τ^∗, at the crossover point between different scaling exponents in the DFA, independent of the time scale at which the fog event is studied. We also show that these scale-invariant characteristic times are related to the duration of fog/no-fog periods, what suggests that τ^∗ is a good measure to study and compare fog events persistence, and also that fog events can be studied as on-off intermittent processes. We apply a binarizing process and confirm that the two-range shape structure in the DFA is maintained in the binary time series associated to fog events. We also show how τ^∗ is related to the binary structure of the time series.

在这项工作中，我们研究了具有不同物理形成机制的事件中雾长期持久性的不同特征。具体来说，我们通过与其相关的低可见度时间序列的去趋势波动分析（DFA），着重从观测数据中对雾的长期持久性进行表征。我们用辐射和地形的潜在物理形成机制分析雾事件，并确定长期持续性的两个范围模式。我们的分析导致特征时间τ ^∗的出现，在DFA中不同缩放指数之间的交叉点，与研究雾事件的时间尺度无关。我们还表明，这些尺度不变的特征时间与有雾/无雾时间段的持续时间有关，这表明τ ^∗是研究和比较雾事件持续性的好方法，而且雾事件可以像断断续续的过程。我们应用了二值化过程，并确认DFA中的两段形状结构在与雾事件相关的二进制时间序列中得以维持。我们还展示了τ ^∗与时间序列的二进制结构之间的关系。