Measurements of the vertical distributions of ozone and related species using high resolution FTIR

我们于1996年10月在名古屋大学茂尻天文台安装了可调谐半导体激光外差光谱仪（TDLHS）。利用TDLHS和傅里叶变换红外光谱仪（FTIR）同时进行了两周的太阳红外光谱测量。对O_3、N_2O、HCl、HF、HNO<@D_23 @> D_2等几种痕量分子进行了吸收光谱测定。此外，还在莫西里天文台进行了臭氧探测仪测量。O3的体积混合比（VMR）的垂直廓线检索这些高分辨率的太阳光谱。VMR剖面反演基于迭代反演算法，该算法利用逐线逐层计算。从40个光谱的两个光谱区检索80臭氧垂直廓线，并与附近站的17个臭氧探空仪测量值进行了比较。18 ~ 34 km的平均差异小于（±）SY. [）12%。18公里和34公里之间的平均差异的标准差小于20%。通过数值模拟研究了初始假设、温度场和仪器线型函数等因素对剖面反演误差的影响。估计的误差是非常一致的观测误差之间的18和34公里，这表明，FTIR的光学良好的调整。在一般情况下，它被发现，检索的O3配置文件提供VMR的准确性优于40%，在14公里，提高到10-15%之间的18和30公里，精度为30%和20%，这些高度范围内，分别。

We installed the Tunable Diode Laser Heterodyne Spectrometer (TDLHS) at Moshiri Observatory, Nagoya University in October, 1996. Measurement of Solar infrared spectra using TDLHS and Fourier-Transform Infrared Spectrometer (FTIR) were made for two weeks simultaneously. Absorption spectra of several trace molecules such as O3, N2O,HCl, HF,and HNO<@D23@>D2 were taken. Also ozonesonde measurements were conducted at Moshiri Observatory. Vertical profiles of volume mixing ratios (VMRs) of O3 were retrieved from these high-resolution solar spectra. The VMR profile retrieval is based on an iterative inversion algorithm, which utilizes line-by-line, layr-by-layr calculations. Eighty ozone vertical profiles were retrieved from the two spectral regions from 40 spectra and were compared with 17 ozonesonde measurements from a nearby station. Average differences between 18 and 34 km were less than (]SY.+-。[) 12%. The standard deviation of the average differences between 18 and 34 km were less than 20%. Effects of several different factors on errors of profile retrievals, such as initial guess, temperature profiles, and instrument line shape functions, were evaluated by numerical simulation. The estimated errors were extremely consistent with the observed errors between 18 and 34 km, suggesting that the FTIR was optically well adjusted. In general, it was found that retrieval of the O3 profile gives VMRs in an accuracy better than 40% at 14 km, improving to 10-15% between 18 and 30 km, with precision of 30% and 20% for these altitude ranges, respectively.