Sensitivity of Bayesian 21 cm power spectrum estimation to foreground model errors
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
ABSTRACT Power spectrum estimators are an important tool in efforts to detect the 21 cm brightness temperature fluctuations from neutral hydrogen at early times. An initial detection will likely be statistical in nature, meaning that it will not be possible to make a coherent map of the brightness temperature fluctuations; instead, only their variance will be measured against a background of noise and residual systematic effects. Optimal quadratic estimator-based methods often apply an inverse covariance weighting to the data. However, inaccurate covariance modelling can lead to reduced sensitivity and, in some cases, severe signal loss. We recently proposed a Bayesian method to jointly estimate the 21 cm fluctuations, their power spectrum, and foreground emission. Instead of requiring a fixed a priori estimate of the covariance, we estimate the covariance as part of the inference. Choices of parametrization, particularly of the foregrounds, are subject to model errors and could lead to biases and other ill effects if not properly controlled. In this paper, we investigate the effects of inaccurate foreground models on 21 cm power spectrum recovery. Using simulated visibilities, we find that, even in the most extreme scenarios tested, our approach is capable of recovering 21 cm delay power spectrum estimates consistent with a known input signal for delays $\gtrsim 300$ ns ($\sim$88 per cent of the available Fourier modes). This is true even when using foreground models derived from modified foreground catalogues containing spatial and spectral perturbations at the quoted level of uncertainty on our foreground catalogues.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it