Zhaoxuan Zhang, Le Zhang, Pengjie Zhang
Abstract This study presents a novel approach to reconstructing the thermal Sunyaev-Zeldovich (tSZ) power spectrum from Planck data using the Analytical Blind Separation (ABS) method. ABS enhances the recovery of weak signals by excluding low-signal-to-noise eigenmodes and introducing a shift parameter to stabilize the reconstruction. Using the Planck PR3 full-mission data, ABS yields lower amplitudes of the y -map power spectrum at ℓ ≳ 300 than the Planck 2015 MILCA and NILC band powers, suggesting reduced contamination from foreground residuals. After marginalizing over residual foregrounds, we estimate the tSZ power spectrum both with and without including the trispectrum contribution in the covariance. Excluding the trispectrum, the inferred tSZ amplitude is ∼ 9% lower than the Planck 2015 best-fit value, yet consistent within 2 σ Astron. Astrophys. 594 (2016) A22. Including the trispectrum yields an amplitude ∼ 18% higher than the “Bolliet 2018” best-fit value, corresponding to a ∼ 1.2 σ deviation Mon. Not. Roy. Astron. Soc. 477 (2018) 4957. In both cases, the results are consistent with the “Battaglia 2012” model at the 1 σ level Astrophys. J. 758 (2012) 74. These results indicate that ABS provides an independent and complementary measurement of the tSZ power spectrum.