The climate change impact appears as a decreasing or increasing monotonic trend in hydro-meteorology time series records due to greenhouse gas (GHG) emissions causing to global warming and climate change impacts. On the other hand, there may be abrupt changes in the form of jumps in these series due to natural and engineering activities. Although trend identification methods are rather common in the literature, jump determination conventional methodologies are rather rare and their applications present some restrictive asumptions like the serial independence and normal (Gaussian) probability distribution function (PDF). The methodology presented in this paper is away from each of such assumptions and it depicts the minimum number of upcrossing along horizontal truncation levels within the time series variation domain. The applications of the methodology are given for annual Danube River discharge records, Romania; New Jersey rainfall and temperature records, USA; monthly rainfall records and Van Lake level fluctuations, Turkey.

由于温室气体（GHG）排放导致全球变暖和气候变化影响，气候变化影响在水文气象时间序列记录中呈现出减少或增加的单调趋势。另一方面，由于自然和工程活动，这些系列的跳跃形式可能会发生突然的变化。尽管趋势识别方法在文献中相当普遍，但跳变确定的常规方法却很少见，其应用也呈现出一些限制性假设，例如序列独立性和正态（高斯）概率分布函数（PDF）。本文提出的方法与上述每个假设都背道而驰，它描述了在时间序列变化域内沿水平截断水平向上交叉的最小数量。该方法的应用已用于罗马尼亚多瑙河的年度排放量记录。美国新泽西州的降雨和温度记录；土耳其的月降雨量记录和Van Lake水位波动。