Analysis of the Long-term Evolution of the Mesopause region derived from OH(3,1) rotational temperatures

长周期的温度变化是大气温度的一个共同特征。一个突出的例子是太阳活动的11年周期，从地面到较低的热层，在许多不同的温度时间序列中都可以观察到这种周期。此外，在地表温度中还观察到了更长周期的温度波动，周期约为20年和约70年。在平流层温度也观察到准双日振荡。在以前的研究中，我们使用GRIPS(地基红外P分支光谱仪)仪器在中层顶区的OH(3，1)旋转温度中也清楚地观察到了这种振荡。这种振荡对中层顶区的任何导出的线性趋势都有很大的影响，即根据所分析的时间间隔，线性趋势有很大的不同，因此非常重要。这种振荡肯定也会影响任何与温度相关的量，如相当于夏季持续时间(ESD)。ESD被定义为温度低于198K的时间间隔。在以往的研究中，ESD随时间的延长而增加。现在对准双年振荡和更长的时间序列的新知识使我们能够对ESD进行新的分析，一方面将ESD的变化归因于长期的温度变化(例如太阳活动的11年周期；准双年振荡)，另一方面则是季节周期形状的变化。在2000年代中期，除温度本身外，其他大气量值也显示出准双周期振荡或趋势突变，即趋势逆转。行星波和重力波的波动活动都属于这些量。因为这些波是平流层和中间层剩余环流的驱动力，所以它们非常重要。已有研究表明，2004年前后行星波活动呈准双日振荡，重力波活动呈趋势突变。但以前的研究在时间上是有限的，因为与现在相比，只有更短的时间序列可用。此外，在重力波活动的情况下，在部分分析期间也存在所考虑的周期范围的限制。利用现在可用的更长的时间序列和改进的技术，可以获得对波活动的长期演变的新的、更详细的分析。这一新技术基于Lomb-Scarger-Period图，可以处理不等间距的时间序列，这对于OH(3，1)旋转温度来说是非常常见的。在以前的研究中，要么限制了分析的时间间隔，从而限制了周期范围(重力波研究)，要么预先内插了数据空白(行星波研究)。因此，我们的新技术将克服这些缺点。

Long-periodic temperature changes are a common feature of atmospheric temperatures. One prominent example is the 11-year cycle of solar activity that is observed in many different temperature time series from ground up to the lower thermosphere. Furthermore, also temperature fluctuations with even longer periods of about 20 years and about 70 years are observed in surface temperatures. The quasi-bidecadal oscillation was also observed in stratospheric temperatures. In previous studies we clearly observed this oscillation also in OH(3,1) rotational temperatures in the mesopause region that have been observed from Wuppertal using GRIPS (GRound-based Infrared P-branch Spectrometers) instruments. This oscillation has a large influence on any derived linear trend in the mesopause region, i.e. depending on the analysed time interval the linear trend largely differs, and is therefore of great importance. The oscillation surely also influences any temperature related quantity such as the equivalent summer duration (ESD). The ESD is defined as the time interval where the temperatures lie below 198 K. In former studies the ESD showed an increase with time. The new knowledge of the quasi-bidecadal oscillation and the much longer time series that is available now enables a new analysis of the ESD and the attribution of ESD changes to long-term temperature changes (e.g. 11-year cycle of solar activity; quasi-bidecadal oscillation) on the one hand and changes of the shape of the seasonal cycle on the other hand. Beside the temperatures themselves also other atmospheric quantities show a quasi-bidecadal oscillation or a trend-break, i.e. a reversal of the trend, in the mid 2000s. The wave activity of the planetary waves and also gravity waves belong to these quantities. Because these waves are the drivers of the residual circulation in the stratosphere and mesosphere they are of great importance. Former studies already showed that the planetary wave activity shows a quasi-bidecadal oscillation and the gravity wave activity shows a trend-break in about 2004. But the former studies were limited in time, as only much shorter time series were available compared to now. Furthermore, in the case of the gravity wave activity also a limitation of the considered period range during parts of the analysis were present. With the longer time series that is available now and with an improved technique a new and more detailed analysis of the long-term evolution of the wave activity can be obtained. The new technique is based on the Lomb-Scargle-Periodogram and can handle time series with unequal spacing which is very common for OH(3,1) rotational temperatures. In the former studies either the analysed time interval and therefore the period range was restricted (gravity waves study) or the data gaps have been interpolated in advance (planetary waves study). Thus, our new technique will overcome these drawbacks.