Extinction curves, extinction laws, and the failure of interstellar dust models

TL;DR

This paper critically examines the interpretation of ultraviolet interstellar extinction curves, proposing a new classification and suggesting that the 2200 A bump may result from scattered starlight rather than traditional dust extinction models.

Contribution

It introduces a new definition of extinction curves, classifies them into three types, and argues that the 2200 A bump is likely due to scattered light rather than dust extinction.

Findings

Extinction curves can be classified into three types.

The 2200 A bump may result from scattered starlight, not dust extinction.

Traditional dust models may not fully explain observed extinction features.

Abstract

The interpretation of ultraviolet Galactic interstellar extinction curves is obscured today by accumulated assumptions, such as a purported link between the 2200 A bump and metallicity, that are not firmly supported by observations. In this paper I define extinction curves as the ratio F*/F0 of the near-infrared-to-ultraviolet spectrum of a reddened star to that of the same star without intervening material, rather than in terms of a magnitude difference, and revisit their observed properties. Special attention is given to the connection that Galactic extinction curves with a 2200 A bump retain with the ultraviolet extrapolation of the exponential extinction law defined by their near-infrared-to-optical segment. This connection leads to the classification of all extinction curves into three types. A graphical representation of these types together with their underlying exponential extinction laws demonstrates that interstellar extinction curves can be interpreted in two ways. Either they result from the mixing of distinct extinction laws associated with different particles, as traditionally assumed, or Galactic ultraviolet curves with a bump are not extinction laws proper but instead deviate from a universal exponential extinction law owing to an additional contribution from coherently forward-scattered starlight. Given the observational constraints on the interpretation of extinction curves, such as their dependence on just two parameters, and the fact that bump-like extinction curves are barely observed outside the Galaxy, the latter interpretation emerges as the only logically consistent one.