Search

For decades, the Free Application for Federal Student Aid (FAFSA) forms were processed electronically by the legacy Central Processing System. In 2021 the Office of Federal Student Aid (FSA) initiated a system development effort and in December 2023 deployed a new system to process forms for the 2024-2025 school year. However, student aid applicants reported that the new system had availability issues, recurring errors, and long wait times. This Government Accountability Office (GAO) testimony (1) describes the delays in delivering on the FAFSA Processing System (FPS) requirements; (2) describes the defects that were identified before and after FPS deployment; (3) assesses the extent to which disciplined systems acquisition practices were used to manage FPS requirements, conduct systems testing, and carry out independent acquisition reviews; and (4) assesses the extent to which Education and FSA IT leadership provided oversight of FPS development. This statement is a companion to GAO's related statement on the simplified FAFSA rollout (GAO-24-107407). In addressing the first two objectives, GAO summarized the preliminary results of its continuing ongoing work describing delays and defects. For the latter two objectives on requirements, testing, and independent reviews, GAO completed its audit work and is making recommendations to address weaknesses.

We explore strategies aimed at reducing the amount of computation, both quantum and classical, required to run the Quantum Approximate Optimization Algorithm (QAOA). First, following Wurtz et al. [Phys.Rev A 104:052419], we consider the standard QAOA with instance-independent "tree" parameters chosen in advance. These tree parameters are chosen to optimize the MaxCut expectation for large girth graphs. We provide extensive numerical evidence supporting the performance guarantee for tree parameters conjectured in [Phys.Rev A 103:042612] and see that the approximation ratios obtained with tree parameters are typically well beyond the conjectured lower bounds, often comparable to performing a full optimization. This suggests that in practice, the QAOA can achieve near-optimal performance without the need for parameter optimization. Next, we modify the warm-start QAOA of Tate et al. [Quantum 7:1121]. The starting state for the QAOA is now an optimized product state associated with a solution of the Goemans-Williamson (GW) algorithm. Surprisingly, the tree parameters continue to perform well for the warm-start QAOA. We find that for random 3-regular graphs with hundreds of vertices, the expected cut obtained by the warm-start QAOA at depth $p \gtrsim 3$ is comparable to that of the standard GW algorithm. Our numerics on random instances do not provide general performance guarantees but do provide substantial evidence that there exists a regime of instance sizes in which the QAOA finds good solutions at low depth without the need for parameter optimization. For each instance studied, we classically compute the expected size of the QAOA distribution of cuts; producing the actual cuts requires running on a quantum computer., Comment: 15 pages, 4 figures

The reduced speed of light approximation has been employed to speed up radiative transfer simulations of reionization by a factor of $\gtrsim 5-10$. However, it has been shown to cause significant errors in the HI-ionizing background near reionization's end in simulations of representative cosmological volumes. This can bias inferences on the galaxy ionizing emissivity required to match observables, such as the Ly$\alpha$ forest. In this work, we show that using a reduced speed of light is, to a good approximation, equivalent to re-scaling the global ionizing emissivity in a redshift-dependent way. We derive this re-scaling and show that it can be used to ``correct'' the emissivity in reduced speed of light simulations. This approach of re-scaling the emissivity after the simulation has been run is useful in contexts where the emissivity is a free parameter. We test our method by running full speed of light simulations using these re-scaled emissivities and comparing them with their reduced speed of light counterparts. We find that for reduced speeds of light $\tilde{c} \geq 0.2$, the 21 cm power spectrum at $0.1 \leq k /[h{\rm Mpc}^{-1}] \leq 0.2$ and key Ly$\alpha$ forest observables agree to within $20\%$ throughout reionization, and often better than $10\%$. Position-dependent time-delay effects cause inaccuracies in reionization's morphology on large scales that produce errors up to a factor of $2$ for $\tilde{c} \leq 0.1$. Our method enables a factor of $5$ speedup of radiative transfer simulations of reionization in situations where the emissivity can be treated as a free parameter., Comment: 23 pages, 12 figures, submitted to JCAP. Comments welcome

The wealth of high-quality observational data from the epoch of reionization that will become available in the next decade motivates further development of modeling techniques for their interpretation. Among the key challenges in modeling reionization are (1) its multi-scale nature, (2) the computational demands of solving the radiative transfer (RT) equation, and (3) the large size of reionization's parameter space. In this paper, we present and validate a new RT code designed to confront these challenges. FlexRT (Flexible Radiative Transfer) combines adaptive ray tracing with a highly flexible treatment of the intergalactic ionizing opacity. This gives the user control over how the intergalactic medium (IGM) is modeled, and provides a way to reduce the computational cost of a FlexRT simulation by orders of magnitude while still accounting for small-scale IGM physics. Alternatively, the user may increase the angular and spatial resolution of the algorithm to run a more traditional reionization simulation. FlexRT has already been used in several contexts, including simulations of the Lyman-$\alpha$ forest of high-$z$ quasars, the redshifted 21cm signal from reionization, as well as in higher resolution reionization simulations in smaller volumes. In this work, we motivate and describe the code, and validate it against a set of standard test problems from the Cosmological Radiative Transfer Comparison Project. We find that FlexRT is in broad agreement with a number of existing RT codes in all of these tests. Lastly, we compare FlexRT to an existing adaptive ray tracing code to validate FlexRT in a cosmological reionization simulation., Comment: 40+7 pages, 24 figures, submitted to JCAP. Comments welcome

Recent JWST observations at $z > 6$ may imply galactic ionizing photon production in excess of prior expectations. Under observationally motivated assumptions about escape fractions, these suggest a $z \sim 8-9$ end to reionization, in strong tension with the $z < 6$ end required by the Ly$\alpha$ forest. In this work, we use radiative transfer simulations to understand what different observations tell us about when reionization ended and when it started. We consider a model that ends too early (at $z \approx 8$) alongside two more realistic scenarios that end late at $z \approx 5$: one that starts late ($z \sim 9$) and another that starts early ($z \sim 13$). We find that the latter requires up to an order-of-magnitude evolution in galaxy ionizing properties at $6 < z < 12$, perhaps in tension with recent measurements of $\xi_{\rm ion}$ by JWST, which indicate little evolution. We also study how these models compare to recent measurements of the Ly$\alpha$ forest opacity, mean free path, IGM thermal history, visibility of $z > 8$ Ly$\alpha$ emitters, and the patchy kSZ signal from the CMB. We find that neither of the late-ending scenarios is conclusively disfavored by any single data set. However, a majority of these observables, spanning several distinct types of observations, prefer a late start. Not all probes agree with this conclusion, hinting at a possible lack of concordance between observables. Observations by multiple experiments (including JWST, Roman, and CMB-S4) in the coming years will either establish a concordance picture of reionization's early stages or reveal systematics in data and/or theoretical modeling., Comment: 21+7 pages, 13+2 figures, submitted to ApJ. Comments welcome

The relationship between Ly$\alpha$ forest opacity and local galaxy density (the opacity-density relation) is a key observational test of late reionization models. Using narrow-band surveys of z=5.7 Ly$\alpha$ emitters centered on quasar sight lines, Christenson et al. (2023) showed that two of the most transmissive forest segments at this redshift intersect galaxy underdensities. This is in tension with models of a strongly fluctuating ionizing background, including some late reionization models, which predict that the vast majority of these segments should intersect overdensities where the ionizing intensity is strongest. We use radiative transfer simulations to explore in detail the opacity-density relation in late reionization models. Fields like the one toward quasar PSO J359-06 -- the more underdense of the two transmissive sight lines in Christenson et al. (2023) -- typically contain recently reionized gas in cosmic voids where the hot temperatures and low densities enhance Ly$\alpha$ transmission. The opacity-density relation's transmissive end is sensitive to the amount of neutral gas in voids, and its morphology, set by the reionization source clustering. These effects are, however, degenerate. We demonstrate that models with very different source clustering can yield similar opacity-density relations when their reionization histories are calibrated to match Ly$\alpha$ forest mean flux measurements at z<6. In models with fixed source clustering, a lower neutral fraction increases the likelihood of intersecting hot, recently reionized gas in voids, increasing the likelihood of observing PSO J359-06. For instance, the probability of observing this field is 15% in a model with neutral fraction $x_{\rm HI}=5\%$ at z=5.7, three times more likely than in a model with $x_{\rm HI}=15\%$. The opacity-density relation may thus provide a complementary probe of reionization's end., Comment: 31 pages, 17 Figures

In this paper we study the indecomposable module categories over $\mathcal{C}(\mathfrak{sl}_N, k)$, the category of integrable level-$k$ respresentations of affine Kac-Moody $\mathfrak{sl}_N$. Our first main result classifies these module categories in the case of generic $k$, i.e. $k$ is sufficiently large relative to $N$. As $\mathcal{C}(\mathfrak{sl}_N, k)$ is a braided tensor category, there is a relative tensor product structure on its category of module categories. In the generic setting we obtain a formula for the relative tensor product rules between the indecomposable module categories. Our second main result classifies the indecomposable module categories over $\mathcal{C}(\mathfrak{sl}_N, k)$ for $N\leq 7$, with no restrictions on $k$. In this non-generic setting, exceptional module categories are obtained. This work relies heavily on previous results by the two authors. In previous literature, module category classification results were known only for $\mathfrak{sl}_2$ and $\mathfrak{sl}_3$., Comment: 37 pages

Automated detection of acoustic signals is crucial for effective monitoring of vocal animals and their habitats across ecologically-relevant spatial and temporal scales. Recent advances in deep learning have made these approaches more accessible. However, there are few deep learning approaches that can be implemented natively in the R programming environment; approaches that run natively in R may be more accessible for ecologists. The "torch for R" ecosystem has made the use of transfer learning with convolutional neural networks accessible for R users. Here, we evaluate a workflow that uses transfer learning for the automated detection of acoustic signals from passive acoustic monitoring (PAM) data. Our specific goals include: 1) present a method for automated detection of gibbon calls from PAM data using the "torch for R" ecosystem; 2) compare the results of transfer learning for six pretrained CNN architectures; and 3) investigate how well the different architectures perform on datasets of the female calls from two different gibbon species: the northern grey gibbon (Hylobates funereus) and the southern yellow-cheeked crested gibbon (Nomascus gabriellae). We found that the highest performing architecture depended on the test dataset. We successfully deployed the top performing model for each gibbon species to investigate spatial of variation in gibbon calling behavior across two grids of autonomous recording units in Danum Valley Conservation Area, Malaysia and Keo Seima Wildlife Sanctuary, Cambodia. The fields of deep learning and automated detection are rapidly evolving, and we provide the methods and datasets as benchmarks for future work.

Fast, reliable logical operations are essential for the realization of useful quantum computers, as they are required to implement practical quantum algorithms at large scale. By redundantly encoding logical qubits into many physical qubits and using syndrome measurements to detect and subsequently correct errors, one can achieve very low logical error rates. However, for most practical quantum error correcting (QEC) codes such as the surface code, it is generally believed that due to syndrome extraction errors, multiple extraction rounds -- on the order of the code distance d -- are required for fault-tolerant computation. Here, we show that contrary to this common belief, fault-tolerant logical operations can be performed with constant time overhead for a broad class of QEC codes, including the surface code with magic state inputs and feed-forward operations, to achieve "algorithmic fault tolerance". Through the combination of transversal operations and novel strategies for correlated decoding, despite only having access to partial syndrome information, we prove that the deviation from the ideal measurement result distribution can be made exponentially small in the code distance. We supplement this proof with circuit-level simulations in a range of relevant settings, demonstrating the fault tolerance and competitive performance of our approach. Our work sheds new light on the theory of fault tolerance, potentially reducing the space-time cost of practical fault-tolerant quantum computation by orders of magnitude.

Long troughs observed in the $z > 5.5$ Ly$\alpha$ and Ly$\beta$ forests are thought to be caused by the last remaining neutral patches during the end phases of reionization -- termed neutral islands. If this is true, then the longest troughs mark locations where we are most likely to observe the reionizing intergalactic medium (IGM). A key feature of the neutral islands is that they are bounded by ionization fronts (I-fronts) which emit Lyman series lines. In this paper, we explore the possibility of directly imaging the outline of neutral islands with a narrowband survey targeting Ly$\alpha$. In a companion paper, we quantified the intensity of I-front Ly$\alpha$ emissions during reionization and its dependence on the spectrum of incident ionizing radiation and I-front speed. Here we apply those results to reionization simulations to model the emissions from neutral islands. We find that neutral islands would appear as diffuse structures that are tens of comoving Mpc across, with surface brightnesses in the range $\approx 1 - 5\times 10^{-21}$ erg s$^{-1}$ cm$^{-2}$ arcsec$^{-2}$. The islands are brighter if the spectrum of ionizing radiation driving the I-fronts is harder, and/or if the I-fronts are moving faster. We develop mock observations for current and futuristic observatories and find that, while extremely challenging, detecting neutral islands is potentially within reach of an ambitious observing program with wide-field narrowband imaging. Our results demonstrate the potentially high impact of low-surface brightness observations for studying reionization., Comment: 24 pages, 10 figures

During reionization, intergalactic ionization fronts (I-fronts) are sources of Ly$\alpha$ line radiation produced by collisional excitation of hydrogen atoms within the fronts. In principle, detecting this emission could provide direct evidence for a reionizing intergalactic medium (IGM). In this paper, we use a suite of high-resolution one-dimensional radiative transfer simulations run on cosmological density fields to quantify the parameter space of I-front Ly$\alpha$ emission. We find that the Ly$\alpha$ production efficiency -- the ratio of emitted Ly$\alpha$ flux to incident ionizing flux driving the front -- depends mainly on the I-front speed and the spectral index of the ionizing radiation. IGM density fluctuations on scales smaller than the typical I-front width produce scatter in the efficiency, but they do not significantly boost its mean value. The Ly$\alpha$ flux emitted by an I-front is largest if 3 conditions are met simultaneously: (1) the incident ionizing flux is large; (2) the incident spectrum is hard, consisting of more energetic photons; (3) the I-front is traveling through a cosmological over-density, which causes it to propagate more slowly. We present a convenient parameterization of the efficiency in terms of I-front speed and incident spectral index. We make these results publicly available as an interpolation table and we provide a simple fitting function for a representative ionizing background spectrum. Our results can be applied as a sub-grid model for I-front Ly$\alpha$ emissions in reionization simulations with spatial and/or temporal resolutions too coarse to resolve I-front structure. In a companion paper, we use our results to explore the possibility of directly imaging Ly$\alpha$ emission around neutral islands during the last phases of reionization., Comment: 21 pages, 6 figures

The ability to efficiently prepare ground states of quantum Hamiltonians via adiabatic protocols is typically limited by the smallest energy gap encountered during the quantum evolution. This presents a key obstacle for quantum simulation and realizations of adiabatic quantum algorithms in large systems, particularly when the adiabatic gap vanishes exponentially with system size. Using QuEra's Aquila programmable quantum simulator based on Rydberg atom arrays, we experimentally demonstrate a method to circumvent such limitations. Specifically, we develop and test a "sweep-quench-sweep" quantum algorithm in which the incorporation of a quench step serves as a remedy to the diverging adiabatic timescale. These quenches introduce a macroscopic reconfiguration between states separated by an extensively large Hamming distance, akin to quantum many-body scars. Our experiments show that this approach significantly outperforms the adiabatic algorithm, illustrating that such quantum quench algorithms can provide a shortcut to adiabaticity for large-scale many-body quantum systems.

Recent quasar absorption line observations suggest that reionization may end as late as $z \approx 5.3$. As a means to search for large neutral hydrogen islands at $z<6$, we revisit long dark gaps in the Ly$\beta$ forest in VLT/X-Shooter and Keck/ESI quasar spectra. We stack the Ly$\alpha$ forest corresponding to both edges of these Ly$\beta$ dark gaps and identify a damping wing-like extended absorption profile. The average redshift of the stacked forest is $z=5.8$. By comparing these observations with reionization simulations, we infer that such a damping wing-like feature can be naturally explained if these gaps are at least partially created by neutral islands. Conversely, simulated dark gaps lacking neutral hydrogen struggle to replicate the observed damping wing features. Furthermore, this damping wing-like profile implies that the volume-averaged neutral hydrogen fraction must be $\langle x_{\rm HI} \rangle \geq 6.1 \pm 3.9\%$ at $z = 5.8$. Our results offer robust evidence that reionization extends below $z=6$., Comment: 8 pages, 5 figures, 1 table; accepted for publication in MNRAS Letters

The COVID-19 pandemic led to a large global effort to sequence SARS-CoV-2 genomes from patient samples to track viral evolution and inform public health response. Millions of SARS-CoV-2 genome sequences have been deposited in global public repositories. The Canadian COVID-19 Genomics Network (CanCOGeN - VirusSeq), a consortium tasked with coordinating expanded sequencing of SARS-CoV-2 genomes across Canada early in the pandemic, created the Canadian VirusSeq Data Portal, with associated data pipelines and procedures, to support these efforts. The goal of VirusSeq was to allow open access to Canadian SARS-CoV-2 genomic sequences and enhanced, standardized contextual data that were unavailable in other repositories and that meet FAIR standards (Findable, Accessible, Interoperable and Reusable). The Portal data submission pipeline contains data quality checking procedures and appropriate acknowledgement of data generators that encourages collaboration. Here we also highlight Duotang, a web platform that presents genomic epidemiology and modeling analyses on circulating and emerging SARS-CoV-2 variants in Canada. Duotang presents dynamic changes in variant composition of SARS-CoV-2 in Canada and by province, estimates variant growth, and displays complementary interactive visualizations, with a text overview of the current situation. The VirusSeq Data Portal and Duotang resources, alongside additional analyses and resources computed from the Portal (COVID-MVP, CoVizu), are all open-source and freely available. Together, they provide an updated picture of SARS-CoV-2 evolution to spur scientific discussions, inform public discourse, and support communication with and within public health authorities. They also serve as a framework for other jurisdictions interested in open, collaborative sequence data sharing and analyses.

The thermal history and structure of the intergalactic medium (IGM) at $z \geq 4$ is an important boundary condition for reionization, and a key input for studies using the Ly$\alpha$ forest to constrain the masses of alternative dark matter candidates. Most such inferences rely on simulations that lack the spatial resolution to fully resolve the hydrodynamic response of IGM filaments and minihalos to HI reionization heating. In this letter, we use high-resolution hydrodynamic+radiative transfer simulations to study how these affect the IGM thermal structure. We find that the adiabatic heating and cooling driven by the expansion of initially cold gas filaments and minihalos sources significant small-scale temperature fluctuations. These likely persist in much of the IGM until $z \leq 4$. Capturing this effect requires resolving the clumping scale of cold, pre-ionized gas, demanding spatial resolutions of $\leq 2$ $h^{-1}$kpc. Pre-heating of the IGM by X-Rays can slightly reduce the effect. Our preliminary estimate of the effect on the Ly$\alpha$ forest finds that, at $\log(k /[{\rm km^{-1} s}]) = -1.0$, the Ly$\alpha$ forest flux power (at fixed mean flux) can increase $\approx 10\%$ going from $8$ and $2$ $h^{-1}$kpc resolution at $z = 4-5$ for gas ionized at $z < 7$. These findings motivate more careful analyses of how the effects studied here affect the Ly$\alpha$ forest., Comment: 6+1 pages, 2+1 figures, accepted for publication in MNRAS letters. Comments welcome

Realizing computationally complex quantum circuits in the presence of noise and imperfections is a challenging task. While fault-tolerant quantum computing provides a route to reducing noise, it requires a large overhead for generic algorithms. Here, we develop and analyze a hardware-efficient, fault-tolerant approach to realizing complex sampling circuits. We co-design the circuits with the appropriate quantum error correcting codes for efficient implementation in a reconfigurable neutral atom array architecture, constituting what we call a fault-tolerant compilation of the sampling algorithm. Specifically, we consider a family of $[[2^D , D, 2]]$ quantum error detecting codes whose transversal and permutation gate set can realize arbitrary degree-$D$ instantaneous quantum polynomial (IQP) circuits. Using native operations of the code and the atom array hardware, we compile a fault-tolerant and fast-scrambling family of such IQP circuits in a hypercube geometry, realized recently in the experiments by Bluvstein et al. [Nature 626, 7997 (2024)]. We develop a theory of second-moment properties of degree-$D$ IQP circuits for analyzing hardness and verification of random sampling by mapping to a statistical mechanics model. We provide evidence that sampling from hypercube IQP circuits is classically hard to simulate and analyze the linear cross-entropy benchmark (XEB) in comparison to the average fidelity. To realize a fully scalable approach, we first show that Bell sampling from degree-$4$ IQP circuits is classically intractable and can be efficiently validated. We further devise new families of $[[O(d^D),D,d]]$ color codes of increasing distance $d$, permitting exponential error suppression for transversal IQP sampling. Our results highlight fault-tolerant compiling as a powerful tool in co-designing algorithms with specific error-correcting codes and realistic hardware., Comment: 27 + 20 pages, 13 Figures

Quantum error correction is believed to be essential for scalable quantum computation, but its implementation is challenging due to its considerable space-time overhead. Motivated by recent experiments demonstrating efficient manipulation of logical qubits using transversal gates (Bluvstein et al., Nature 626, 58-65 (2024)), we show that the performance of logical algorithms can be substantially improved by decoding the qubits jointly to account for physical error propagation during transversal entangling gates. We find that such correlated decoding improves the performance of both Clifford and non-Clifford transversal entangling gates, and explore two decoders offering different computational runtimes and accuracies. By considering deep logical Clifford circuits, we find that correlated decoding can significantly improve the space-time cost by reducing the number of rounds of noisy syndrome extraction per gate. These results demonstrate that correlated decoding provides a major advantage in early fault-tolerant computation, and indicate it has considerable potential to reduce the space-time cost in large-scale logical algorithms., Comment: 7+12 pages, 5+3 figures

Detection of the faint 21 cm line emission from the Cosmic Dawn and Epoch of Reionisation will require not only exquisite control over instrumental calibration and systematics to achieve the necessary dynamic range of observations but also validation of analysis techniques to demonstrate their statistical properties and signal loss characteristics. A key ingredient in achieving this is the ability to perform high-fidelity simulations of the kinds of data that are produced by the large, many-element, radio interferometric arrays that have been purpose-built for these studies. The large scale of these arrays presents a computational challenge, as one must simulate a detailed sky and instrumental model across many hundreds of frequency channels, thousands of time samples, and tens of thousands of baselines for arrays with hundreds of antennas. In this paper, we present a fast matrix-based method for simulating radio interferometric measurements (visibilities) at the necessary scale. We achieve this through judicious use of primary beam interpolation, fast approximations for coordinate transforms, and a vectorised outer product to expand per-antenna quantities to per-baseline visibilities, coupled with standard parallelisation techniques. We validate the results of this method, implemented in the publicly-available matvis code, against a high-precision reference simulator, and explore its computational scaling on a variety of problems., Comment: 25 pages, 20 figures, submitted to RAS Techniques and Instruments, matvis is publicly available at https://github.com/HERA-Team/matvis

Suppressing errors is the central challenge for useful quantum computing, requiring quantum error correction for large-scale processing. However, the overhead in the realization of error-corrected ``logical'' qubits, where information is encoded across many physical qubits for redundancy, poses significant challenges to large-scale logical quantum computing. Here we report the realization of a programmable quantum processor based on encoded logical qubits operating with up to 280 physical qubits. Utilizing logical-level control and a zoned architecture in reconfigurable neutral atom arrays, our system combines high two-qubit gate fidelities, arbitrary connectivity, as well as fully programmable single-qubit rotations and mid-circuit readout. Operating this logical processor with various types of encodings, we demonstrate improvement of a two-qubit logic gate by scaling surface code distance from d=3 to d=7, preparation of color code qubits with break-even fidelities, fault-tolerant creation of logical GHZ states and feedforward entanglement teleportation, as well as operation of 40 color code qubits. Finally, using three-dimensional [[8,3,2]] code blocks, we realize computationally complex sampling circuits with up to 48 logical qubits entangled with hypercube connectivity with 228 logical two-qubit gates and 48 logical CCZ gates. We find that this logical encoding substantially improves algorithmic performance with error detection, outperforming physical qubit fidelities at both cross-entropy benchmarking and quantum simulations of fast scrambling. These results herald the advent of early error-corrected quantum computation and chart a path toward large-scale logical processors., Comment: See ancillary files: five supplementary movies and captions. Main text + Methods

Observations with radio arrays that target the 21-cm signal originating from the early Universe suffer from a variety of systematic effects. An important class of these are reflections and spurious couplings between antennas. We apply a Hamiltonian Monte Carlo sampler to the modelling and mitigation of these systematics in simulated Hydrogen Epoch of Reionisation Array (HERA) data. This method allows us to form statistical uncertainty estimates for both our models and the recovered visibilities, which is an important ingredient in establishing robust upper limits on the Epoch of Reionisation (EoR) power spectrum. In cases where the noise is large compared to the EoR signal, this approach can constrain the systematics well enough to mitigate them down to the noise level for both systematics studied. Where the noise is smaller than the EoR, our modelling can mitigate the majority of the reflections with there being only a minor level of residual systematics, while cross-coupling sees essentially complete mitigation. Our approach performs similarly to existing filtering/fitting techniques used in the HERA pipeline, but with the added benefit of rigorously propagating uncertainties. In all cases it does not significantly attenuate the underlying signal., Comment: 19 pages, 14 figures, submitted to MNRAS

Quasar absorption spectra measurements suggest that reionization proceeded rapidly, ended late at $z \sim 5.5$, and was followed by a flat evolution of the ionizing background. Simulations that can reproduce this behavior often rely on a fine-tuned galaxy ionizing emissivity, which peaks at $z \sim 6 - 7$ and drops by a factor of $1.5-2.5$ by $z \sim 5$. This is puzzling since the abundance of galaxies has been observed to grow monotonically during this period. Explanations for this include effects such as dust obscuration of ionizing photon escape and feedback due to photo-heating of the IGM. We explore the possibility that this drop in emissivity is instead an artifact of one or more modeling deficiencies in reionization simulations. These include possibly incorrect assumptions about the ionizing spectrum and/or inaccurate modeling of the clumpiness of the IGM. Our results suggest that the need for a drop could be alleviated if simulations are underestimating the IGM opacity from massive, star-forming halos. Other potential modeling issues either have a small effect or require a steeper drop when remedied. We construct an illustrative model in which the emissivity is nearly flat at the end of reionization, evolving only $\sim 0.05$ dex at $5 < z < 7$. More realistic scenarios, however, require a $\sim 0.1-0.3$ dex drop. We also study the evolution of the Ly$\alpha$ effective optical depth distribution in these scenarios and compare them to recent measurements. We find models that feature a hard ionizing spectrum and/or are driven by faint, low-bias sources can most easily reproduce the mean transmission and optical depth distribution of the forest simultaneously. Lastly, we show that the reduced speed of light approximation and low spatial resolution in the forest can lead to erroneous conclusions about the end of reionization., Comment: 20 pages, 15 figures, to be submitted to MNRAS. Comments welcome

Recent measurements of the ionizing photon mean free path (MFP) based on composite quasar spectra may point to reionization ending at $z<6$. These measurements are challenging because they rely on assumptions about the proximity zones of the quasars. For example, some quasars might have been close to neutral patches where reionization was still ongoing ("neutral islands"), and it is unclear how they would affect the measurements. We address this question with mock MFP measurements from radiative transfer simulations. We find that, even in the presence of neutral islands, our mock MFP measurements agree to within $30~\%$ with the true spatially averaged MFP in our simulations, which includes opacity from both the ionized medium and the islands. The inferred MFP is sensitive at the $<~50\%$ level to assumptions about quasar environments and lifetimes for realistic models. We demonstrate that future analyses with improved data may require explicitly modeling the effects of neutral islands on the composite spectra, and we outline a method for doing this. Lastly, we quantify the effects of neutral islands on Lyman-series transmission, which has been modeled with optically thin simulations in previous MFP analyses. Neutral islands can suppress transmission at $\lambda_{\rm rest}<912$ Angstroms significantly, up to a factor of 2 for $z_{\rm qso}=6$ in a plausible reionization scenario, owing to absorption by many closely spaced lines as quasar light redshifts into resonance. However, the suppression is almost entirely degenerate with the spectrum normalization and thus does not significantly bias the inferred MFP., Comment: Accepted to MNRAS, minor changes