AI 中文总结
本研究提出局部带功率诊断方法，检验CMB E模式极化的半球功率不对称性起源，通过Planck数据验证信号的CMB层面属性，为后续LiteBIRD应用奠定基础。
AI 中文摘要
我们提出了局部带功率诊断方法，用于检验E模式极化数据中的候选偶极调制是否源自宇宙微波背景（CMB）本身，而非噪声或系统误差。研究目标是半球功率不对称性（HPA），它是CMB温度场中最持久的大尺度异常之一；极化为其起源提供了独立的探测手段，但在Planck卫星的灵敏度下，极化天空以噪声为主。我们在被掩蔽天空的重叠斑块中估计局部功率，通过标准伪$C_\u2113$反卷积校正E到B的泄漏，并分别对两幅共享同一天空但噪声独立的Planck半任务图拟合偶极方向。方向的高度一致性表明存在CMB层面的信号，尽管无法区分真实调制与偶然的CMB涨落。将该方法应用于$\u2113=15$-40频段的Planck PR3 Commander数据时，两个半任务的方向间隔为$8.7^\u00b0$，比300个各向同性模拟中的任何一个都更一致（$p<0.33\%$），且位于温度HPA轴附近。这种相干性仅出现在EE模式中，未经泄漏校正时会消失，并通过交叉谱检验得到证实。极化灵敏度大幅提升的LiteBIRD是自然的下一步应用对象。
英文摘要
We present local bandpower diagnostics that test whether a candidate dipolar modulation in E-mode polarization data originates in the cosmic microwave background (CMB) itself rather than in noise or systematics. The target is the hemispherical power asymmetry (HPA), one of the most persistent large-scale anomalies in the CMB temperature field; polarization offers an independent probe of its origin, but at Planck sensitivity the polarized sky is noise-dominated. Local power is estimated in overlapping patches on the masked sky, with E-to-B leakage corrected through standard pseudo-$C_\ell$ deconvolution, and dipole directions are fitted independently to the two Planck half-mission maps, which share the same sky but carry independent noise. Tight directional agreement indicates a CMB-level signal, although it cannot distinguish a real modulation from a chance CMB fluctuation. Applied to Planck PR3 Commander data at $\ell = 15$-$40$, the half-mission directions are separated by $8.7^\circ$, a closer agreement than in any of the 300 isotropic simulations ($p < 0.33\%$), and lie near the temperature HPA axis. The coherence appears only in EE, vanishes without leakage correction, and is corroborated by a cross-spectrum check. LiteBIRD, with substantially higher polarization sensitivity, is a natural next application.
Journal refMNRAS 550 (2026) stag1312
