Fitting the curve in Excel®: Systematic curve fitting of laboratory and remotely sensed planetary spectra. (March 2017)

• Record Type: Journal Article
• Authors: McCraig, Michael A.; Osinski, Gordon R.; Cloutis, Edward A.; Flemming, Roberta L.; Izawa, Matthew R.M.; Reddy, Vishnu; Fieber-Beyer, Sherry K.; Pompilio, Loredana; van der Meer, Freek; Berger, Jeffrey A.; Bramble, Michael S.; Applin, Daniel M.
• Is Part Of: Computers & geosciences. Volume 100(2017)
• Page Start: 103
• Page End: 114
• Publication Date: 2017-03
• Subjects: Data reduction techniques -- Curve fitting -- Spectroscopy -- Spectral analysis -- Curve deconvolution -- Modelling -- Asteroids -- Mineralogy -- Mars; Environmental policy -- Periodicals; 550.5
• DOI: 10.1016/j.cageo.2016.11.018
• Languages: English
• ISSNs: 0098-3004

Abstract

Abstract: Spectroscopy in planetary science often provides the only information regarding the compositional and mineralogical make up of planetary surfaces. The methods employed when curve fitting and modelling spectra can be confusing and difficult to visualize and comprehend. Researchers who are new to working with spectra may find inadequate help or documentation in the scientific literature or in the software packages available for curve fitting. This problem also extends to the parameterization of spectra and the dissemination of derived metrics. Often, when derived metrics are reported, such as band centres, the discussion of exactly how the metrics were derived, or if there was any systematic curve fitting performed, is not included. Herein we provide both recommendations and methods for curve fitting and explanations of the terms and methods used. Techniques to curve fit spectral data of various types are demonstrated using simple-to-understand mathematics and equations written to be used in Microsoft Excel® software, free of macros, in a cut-and-paste fashion that allows one to curve fit spectra in a reasonably user-friendly manner. The procedures use empirical curve fitting, include visualizations, and ameliorates many of the unknowns one may encounter when using black-box commercial software. The provided framework is a comprehensive record of the curve fitting parameters used, the derived metrics, and is intended to be an example of a format for dissemination when curve fitting data. Highlights: An introduction to empirical curve fitting UV–vis–NIR, and other spectra. An introduction to semi-automated methods for Gaussian Fitting. An introduction to common data analysis techniques used by reflectance spectroscopists. Curve fitting using Microsoft Excel.