A novel method to detect the tropopause structure based on bi-Gaussian function

Abstract. The tropopause is important as a diagnostic of the upper troposphere and lower stratosphere structures, with unique atmospheric thermal, dynamic structures. A comprehensive understanding of the evolution of the fine tropopause structures is necessary and primary to further study the complex multi-scale atmospheric physicochemical coupling processes in the upper troposphere and lower stratosphere. Utilizing the bi-Gaussian function, a novel method is capable of identifying the characteristic parameters of tropopause vertical structures, as well as providing the information of double tropopauses (DT) structures. The new method improves the definition of cold point tropopause, and detects one (or two) most significant local coldest point(s) in mathematical statistics by fitting the temperature profiles to the bi-Gaussian function, which is (are) defined as the tropopause height(s). The bi-Gaussian function exhibits remarkable potential for explicating the variation trend of temperature profiles. The recognition results of the bi-Gaussian method and lapse rate tropopause, as defined by World Meteorological Organization, are compared in detail for different cases. Results indicate that the bi-Gaussian method possesses a lower missed detection rate and false detection rate than lapse rate tropopause, because it is not restricted by thresholds, even in the presence of multiple temperature inversion layers at higher elevations. Five-year (from 2012 to 2016) historical radiosondes in China revealed that the occurrence frequency and thickness of DT, as well as the single tropopause height, and the first and second DT height displayed significant meridional monotonic variations. The occurrence frequency (thickness) of DT increased from 2.93 % (2.61 km) to 72.45 % (6.84 km) in the latitude range [16° N, 50° N]. At mid-latitudes [30° N, 40° N], the meridional gradients of tropopause height were relatively large, essentially corresponding to the climatological location of the subtropical jet and Tibetan Plateau. The average DT thickness reported in this study is approximately 1–2 km thicker than that in previous studies, particularly in the mid-high latitudes [45° N, 50° N], which may be related to the different vertical resolution of temperature profiles provided by various data sources. DT structure occurs most frequently and has the largest meridional gradient in the mid-latitudes, formatted by a combination of poleward advection in the low-latitude upper troposphere and equatorward advection in the high-latitude lower stratosphere. In addition, although DT is thick in winter, the DT temperature difference is small, even the case of the first tropopause temperature is lower than the second tropopause temperature happens occasionally.