Several past studies have described how absorption spectroscopy can be used to determine spatial temperature variations along the optical path by measuring the unique, nonlinear response to temperature of many molecular absorption transitions and performing an inversion. New laser absorption spectroscopy techniques are well-suited to this nonuniformity measurement, yet present analysis approaches use only isolated features rather than a full broadband spectral measurement. In this work, we develop a constrained spectral fitting technique called E${}^{″}$-binning to fit an absorption spectrum arising from a nonuniform environment. The information extracted from E${}^{″}$-binning is then input to the inversion approach from the previous paper in this series (Malarich and Rieker, JQSRT 107455 [1]) to determine the temperature distribution. We demonstrate this approach by using dual frequency comb laser measurements to resolve convection cells in a tube furnace. The recovered temperature distributions at each measurement height agree with an existing natural convection model. Finally, we show that for real-world measurements with noise and absorption model error, increasing the bandwidth and the number of measured absorption transitions may improve the temperature distribution precision. We make the fitting code publicly available for use with any broadband absorption measurement.

过去的几项研究描述了如何使用吸收光谱法通过测量许多分子吸收跃迁对温度的独特非线性响应并执行反演来确定沿光路的空间温度变化。新的激光吸收光谱技术非常适合这种不均匀性测量，但目前的分析方法仅使用孤立的特征，而不是全宽带光谱测量。在这项工作中，我们开发了一种称为 E 的约束光谱拟合技术${}^{”}$-binning 以适应非均匀环境产生的吸收光谱。从 E 中提取的信息${}^{”}$然后将 -binning 输入到本系列前一篇论文（Malarich 和 Rieker，JQSRT 107455 [1]）中的反演方法中，以确定温度分布。我们通过使用双频梳状激光测量来解决管式炉中的对流单元来演示这种方法。每个测量高度的恢复温度分布与现有的自然对流模型一致。最后，我们表明，对于具有噪声和吸收模型误差的实际测量，增加带宽和测量吸收跃迁的数量可能会提高温度分布精度。我们公开拟合代码以用于任何宽带吸收测量。