Accelerated inference on accelerated cosmic expansion: New constraints on axionlike early dark energy with DESI BAO and ACT DR6 CMB lensing
Frank J. Qu, Kristen M. Surrao, Boris Bolliet, J. Colin Hill, Blake D. Sherwin, Hidde T. Jense, Adrien La Posta
Abstract
The early dark energy (EDE) extension to <a:math xmlns:a="http://www.w3.org/1998/Math/MathML" display="inline"> <a:mi mathvariant="normal">Λ</a:mi> </a:math> cold dark matter ( <d:math xmlns:d="http://www.w3.org/1998/Math/MathML" display="inline"> <d:mi mathvariant="normal">Λ</d:mi> <d:mi>CDM</d:mi> </d:math> ) has been proposed as a candidate scenario to resolve the “Hubble tension.” We present new constraints on the EDE model by incorporating new data from the Dark Energy Spectroscopic Instrument (DESI) baryon acoustic oscillation (BAO) survey and cosmic microwave background (CMB) lensing measurements from the Atacama Cosmology Telescope (ACT) sixth data release and Planck NPIPE data. We do not find evidence for EDE. The maximum fractional contribution of EDE to the total energy density is <g:math xmlns:g="http://www.w3.org/1998/Math/MathML" display="inline"> <g:msub> <g:mi>f</g:mi> <g:mi>EDE</g:mi> </g:msub> <g:mo><</g:mo> <g:mn>0.091</g:mn> </g:math> [95% confidence level (CL)] from our baseline combination of Planck CMB, CMB lensing, and DESI BAO. Our strongest constraints on EDE come from the combination of Planck CMB and CMB lensing alone, yielding <i:math xmlns:i="http://www.w3.org/1998/Math/MathML" display="inline"> <i:msub> <i:mi>f</i:mi> <i:mi>EDE</i:mi> </i:msub> <i:mo><</i:mo> <i:mn>0.070</i:mn> <i:mo stretchy="false">(</i:mo> <i:mn>95</i:mn> <i:mo>%</i:mo> <i:mi>CL</i:mi> <i:mo stretchy="false">)</i:mo> </i:math> . We also explore extensions of <m:math xmlns:m="http://www.w3.org/1998/Math/MathML" display="inline"> <m:mi mathvariant="normal">Λ</m:mi> <m:mi>CDM</m:mi> </m:math> beyond the EDE parameters by treating the total neutrino mass as a free parameter, finding <p:math xmlns:p="http://www.w3.org/1998/Math/MathML" display="inline"> <p:mo>∑</p:mo> <p:msub> <p:mi>m</p:mi> <p:mi>ν</p:mi> </p:msub> <p:mo><</p:mo> <p:mn>0.096</p:mn> <p:mtext> </p:mtext> <p:mtext> </p:mtext> <p:mi>eV</p:mi> <p:mo stretchy="false">(</p:mo> <p:mn>95</p:mn> <p:mo>%</p:mo> <p:mi>CL</p:mi> <p:mo stretchy="false">)</p:mo> </p:math> and <t:math xmlns:t="http://www.w3.org/1998/Math/MathML" display="inline"> <t:msub> <t:mi>f</t:mi> <t:mi>EDE</t:mi> </t:msub> <t:mo><</t:mo> <t:mn>0.087</t:mn> <t:mo stretchy="false">(</t:mo> <t:mn>95</t:mn> <t:mo>%</t:mo> <t:mi>CL</t:mi> <t:mo stretchy="false">)</t:mo> </t:math> . For the first time in EDE analyses, we perform Bayesian parameter estimation using neural network emulators of cosmological observables, which are on the order of 100 times faster than full Boltzmann solutions.