A Statistically Robust 3-Sigma Detection of Non-Gaussianity in the WMAP Data Using Hot and Cold Spots

TL;DR

This paper introduces a statistically robust frequentist method to detect non-Gaussian features in WMAP CMB data, identifying a new 3-sigma anomaly related to hot and cold spots.

Contribution

It presents a novel hypothesis testing approach that reliably detects non-Gaussianity in CMB data with controlled false detection risk.

Findings

Detection of a 3-sigma non-Gaussian anomaly in WMAP data

Method demonstrates robustness under different noise and resolution models

Software implementation of the test is publicly available

Abstract

We present a careful frequentist analysis of one- and two-point statistics of the hot and cold spots in the cosmic microwave background (CMB) data obtained by the Wilkinson Microwave Anisotropy Probe (WMAP). Our main result is the detection of a new anomaly at the 3-sigma level using temperature-weighted extrema correlation functions. We obtain this result using a simple hypothesis test which reduces the maximum risk of a false detection to the same level as the claimed significance of the test. We further present a detailed study of the robustness of our earlier claim (Larson and Wandelt 2004) under variations in the noise model and in the resolution of the map. Free software which implements our test is available online.