# Broad absorption line disappearance and emergence using multiple-epoch spectroscopy from the Sloan Digital Sky Survey

## TL;DR

This study analyzes the variability of broad absorption lines in quasars over several years, revealing their lifetimes, possible physical mechanisms, and spatial locations, using multi-epoch SDSS spectroscopy of 470 BAL quasars.

## Contribution

It provides the first detailed statistics on BAL disappearance and emergence over multi-year timescales, and constrains the physical locations and sizes of BAL outflows.

## Key findings

- Disappearing BALs occur at a rate of 2.3% per year.
- Emerging BALs occur at a rate of 3.0% per year.
- BAL lifetimes are estimated to be less than 100-1000 years.

## Abstract

We investigate broad absorption line (BAL) disappearance and emergence using a 470 BAL-quasar sample over < 0.10-5.25 rest-frame years with at least three spectroscopic epochs for each quasar from the Sloan Digital Sky Survey. We identify 14 disappearing BALs over < 1.73-4.62 rest-frame years and 18 emerging BALs over < 1.46-3.66 rest-frame years associated with the CIV 1548,1550 and/or SiIV 1393,1402 doublets, and report on their variability behavior. BAL quasars in our dataset exhibit disappearing/emerging CIV BALs at a rate of 2.3 and 3.0 per cent, respectively, and the frequency for BAL to non-BAL quasar transitions is 1.7 per cent. We detect four re-emerging BALs over < 3.88 rest-frame years on average and three re-disappearing BALs over < 4.15 rest-frame years on average, the first reported cases of these types. We infer BAL lifetimes along the line of sight to be nominally < 100-1000 yr using disappearing CIV BALs in our sample. Interpretations of (re-)emerging and (re-)disappearing BALs reveal evidence that collectively supports both transverse-motion and ionization-change scenarios to explain BAL variations. We constrain a nominal CIV/SiIV BAL outflow location of < 100 pc from the central source and a radial size of > 1x10^-7 pc (0.02 au) using the ionization-change scenario, and constrain a nominal outflow location of < 0.5 pc and a transverse size of ~0.01 pc using the transverse-motion scenario. Our findings are consistent with previous work, and provide evidence in support of BALs tracing compact flow geometries with small filling factors.