Mapping the Cosmic Microwave Background Anisotropy:Combined Analysis of QMAP Flights

A. de Oliveira‐Costa, Michelle Devlin, Tom Herbig, Amber Miller, C. B. Netterfield, Lyman A. Page, Max TegmarkThe Astrophysical Journal 1998 · DOI · arXiv

Abstract

We present results from the QMAP balloon experiment, which maps the Cosmic Microwave Background (CMB) and probes its angular power spectrum on degree scales. In two separate flights, data were taken in six channels at two frequency bands between 26 to 46 GHz. We describe our method for mapmaking (removal of 1/f-noise and scan-synchronous offsets) and power spectrum estimation, as well as the results of a joint analysis of the data from both flights. This produces a 527 square degree map of the CMB around the North Celestial Pole, allowing a wide variety of systematic cross-checks. The frequency dependence of the fluctuations is consistent with CMB and inconsistent with Galactic foreground emission. The anisotropy is measured in three multipole bands from l~40 to l~200, and the angular power spectrum shows a distinct rise which is consistent with the Saskatoon results.

Introduction

In the late 1990s, the field of cosmology was focused on characterizing the tiny temperature fluctuations in the Cosmic Microwave Background (CMB)—the "seeds" of all cosmic structure. While the COBE satellite had successfully mapped these fluctuations on very large scales (low multipole $\ell$), a significant gap remained at "degree scales" ($\ell \approx 40–200$). Probing these scales is essential for testing inflationary models and determining fundamental cosmological parameters.

However, balloon-borne experiments like QMAP faced significant technical challenges. Specifically, instrumental $1/f$ noise (long-term drifts) and scan-synchronous offsets (thermal emissions from the telescope itself) can easily overwhelm the faint CMB signal. This paper details the rigorous mathematical and observational strategies used to overcome these hurdles and produce a reliable map of the early universe.

Methodology and Data

The QMAP experiment utilized two independent balloon flights to ensure redundancy and allow for cross-validation:

  • Flight 1 (FL1): Launched June 1996 from Palestine, Texas.
  • Flight 2 (FL2): Launched November 1996 from Ft. Sumner, New Mexico.

Data were collected across six channels in the Ka-band (~30 GHz) and Q-band (~40 GHz). The survey targeted the North Celestial Pole, intentionally overlapping with the previous Saskatoon experiment.

Mapmaking as a Linear Algebra Problem

The researchers modeled the raw data ($y$) as a combination of the true sky temperature ($x$), the scan strategy ($A$), and noise ($n$): $y = Ax + n$. To solve this efficiently, they introduced several key innovations:

  1. Pre-blueing Filters: To handle the computationally expensive noise covariance matrix, a high-pass circulant filter was applied to "whiten" the noise, making the problem solvable on standard workstations.
  2. Virtual Pixels: To account for instrumental offsets tied to the telescope's scan position, 160 "virtual pixels" were added to the model. This allowed the software to isolate and subtract internal noise from the celestial signal.
  3. Wiener Filtering: This technique was used to optimally suppress noise in the final visual representations, particularly in regions with lower signal-to-noise ratios.

Results: The 527-Square-Degree Map

The combined analysis of both flights resulted in a high-fidelity map covering 527 square degrees of the sky. The signal was detected at a remarkable 15$\sigma$ level, effectively ruling out random noise as the source of the observed fluctuations.

The 527-square-degree QMAP temperature fluctuation map centered on the North Celestial Pole.

Crucially, the frequency dependence of the fluctuations (comparing Ka and Q bands) yielded a spectral index $\beta > -1.4$. This confirmed that the signal was indeed the CMB's blackbody radiation rather than Galactic foregrounds like synchrotron emission or dust.

The Angular Power Spectrum: Confirming the Rise

The primary scientific goal was to measure the angular power spectrum. The QMAP data revealed a distinct increase in power as the scale moved from $\ell \sim 40$ toward $\ell \sim 200$.

The angular power spectrum of the CMB, showing QMAP data points (Ka and Q bands) alongside COBE and Saskatoon (SK) results.

This "rise" in the power spectrum was a pivotal observation. It provided strong evidence for the first "acoustic peak," a fundamental prediction of standard Big Bang cosmology. The temperature fluctuations ($\delta T$) were measured in the range of 47–59 $\mu$K, showing striking consistency with the earlier Saskatoon results.

Discussion and Limitations

While the QMAP results were robust, the analysis involved trade-offs. The aggressive high-pass filtering required to remove instrumental offsets also removed the CMB dipole, limiting the experiment's ability to probe the very largest angular scales. Additionally, while statistical errors were minimized, absolute calibration uncertainties remained a factor for future refinement.

Despite these limitations, the QMAP experiment demonstrated that sophisticated filtering and joint-flight analysis could produce high-precision maps of the CMB, bridging the gap between large-scale satellite missions and small-scale ground-based observations.

Summary updated

Abstract

We present results from the QMAP balloon experiment, which maps the Cosmic Microwave Background (CMB) and probes its angular power spectrum on degree scales. In two separate flights, data were taken in six channels at two frequency bands between 26 to 46 GHz. We describe our method for mapmaking (removal of 1/f-noise and scan-synchronous offsets) and power spectrum estimation, as well as the results of a joint analysis of the data from both flights. This produces a 527 square degree map of the CMB around the North Celestial Pole, allowing a wide variety of systematic cross-checks. The frequency dependence of the fluctuations is consistent with CMB and inconsistent with Galactic foreground emission. The anisotropy is measured in three multipole bands from l~40 to l~200, and the angular power spectrum shows a distinct rise which is consistent with the Saskatoon results.
BibTeX
@article{oliveiracosta1998mapping,
  author = {A. de Oliveira‐Costa and Michelle Devlin and Tom Herbig and Amber Miller and C. B. Netterfield and Lyman A. Page and Max Tegmark},
  title = {{Mapping the Cosmic Microwave Background Anisotropy:Combined Analysis of QMAP Flights}},
  journal = {The Astrophysical Journal},
  year = {1998},
  doi = {10.1086/311767},
}
Figure