🔭 astrophysics

Dish Assembly Precision for HIRAX

This paper presents photogrammetry-based measurements of the surface and feed placement precision for the first 28 HIRAX radio telescope dishes, establishing a critical baseline for modeling and mitigating instrumental systematics that could otherwise leak foreground signals into the cosmological hydrogen intensity mapping data.

Original authors: Jennifer Studer, Devin Crichton, Alexandre Refregier, Keshav Bechoo, H. Cynthia Chiang, Chandrachud B. V. Dash, Sindhu Gaddam, Aditya Krishna Karigiri Madhusudhan, Martin Kunz, Kavilan Moodley, Warren

Published 2026-07-30

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Jennifer Studer, Devin Crichton, Alexandre Refregier, Keshav Bechoo, H. Cynthia Chiang, Chandrachud B. V. Dash, Sindhu Gaddam, Aditya Krishna Karigiri Madhusudhan, Martin Kunz, Kavilan Moodley, Warren Naidoo, Tasmiya Papiah, Isibabale Qhoboshiyane, Liantsoa F. Randrianjanahary, Benjamin R. B. Saliwanchik, Ajith Sampath, Corrie Ungerer, Thierry Viant, Anthony Walters

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, invisible ocean, but instead of water, it's filled with hydrogen gas—the same stuff that makes up the stars and your own body. Astronomers want to map this ocean to understand how the universe is stretching and growing, a process driven by a mysterious force called "dark energy." To do this, they use a special kind of telescope that listens for a faint, whisper-like radio signal coming from that hydrogen gas. This signal is so quiet that it's like trying to hear a single pin drop in the middle of a roaring stadium full of cheering fans. The "fans" are bright, natural radio signals from our own galaxy and other galaxies, which are millions of times louder than the hydrogen whisper.

To catch this whisper, scientists are building a massive array of radio dishes, looking like a field of giant satellite TV dishes, all working together as one super-telescope. The trick is that all these dishes need to be almost perfectly identical. If even one dish is slightly bent or if its antenna is placed a tiny bit off-center, the "loud" noise from the stadium can leak through and drown out the "pin drop" signal. This paper is about checking the quality control of the first batch of these dishes to make sure they are built with the precision needed to hear the universe's secrets.

The Great Dish Check-Up

The Hydrogen Intensity and Real-time Analysis eXperiment (HIRAX) is a massive radio telescope array currently being built in the Karoo desert of South Africa. Think of HIRAX as a team of 6-meter-wide (about 20 feet) fiberglass satellite dishes, all standing in a field, working together to map the southern sky. Their job is to catch the faint radio whispers of neutral hydrogen (HI) from billions of years ago. But here's the catch: to do this, the dishes need to be incredibly precise. If the surface of a dish is bumpy or if the antenna (the "feed") holding the receiver is even a little bit crooked, the telescope gets confused. It lets the bright, noisy foreground signals leak into the quiet zone where the cosmological signal lives, ruining the data.

This paper is essentially a report card for the first 28 dishes built for the project. The team used a high-tech camera system called photogrammetry—basically taking hundreds of photos from different angles to build a 3D model of the dishes—to measure two critical things: how smooth the dish surface is and exactly where the feed antenna is sitting.

The Surface: Smooth as Glass or Bumpy as a Potato?

The dishes are made of fiberglass with a hidden aluminum mesh inside that acts as the mirror. The goal was for the surface to be so smooth that any bumps were smaller than 1.0 millimeter. When the team measured the first 28 dishes in the factory, they found something interesting. Most of the dishes were great, but they came from three different molds (think of them as three different cookie cutters).

Two of the molds (M1 and M2) produced dishes that were nearly identical to each other. However, the third mold (M3) had a slight flaw; it made dishes that were about 3 millimeters deeper than the others. Because of this, the team had to look at the data in two ways. If they treated all 28 dishes as one big group, only 36% of them met the strict "precision" requirement (meaning they were all close enough to each other). But, if they separated the M3 dishes from the M1 and M2 dishes, 96% of the total dishes passed the precision test!

The "accuracy" (how close the average dish is to the perfect design) was good for the M1 and M2 group, but the M3 dishes were consistently off by a bit more. The paper suggests that for the final telescope, the team might need to treat the M3 dishes as a slightly different "sub-group" when doing their math, rather than trying to force them to fit the same mold as the others.

The Feed: The Antenna's Seat

Next, the team checked the "feed"—the little antenna held up by four legs right in the center of the dish. This needs to be placed with extreme precision. The rules say the antenna should be within 0.5 millimeters of its perfect spot (precision) and within 3.0 millimeters of the average spot (accuracy).

The results here were a mix of good news and a "we need to work on this" warning. The team found that the accuracy was perfect: on average, the antennas were exactly where they needed to be. However, the precision was a bit shaky. While the angle of the antenna (how much it was tilted) was spot-on, the side-to-side and front-to-back positions varied more than the strict 0.5 mm limit allowed.

Why the wobble? The paper explains that the team uses the measurement tools as a "feedback loop." If a dish is built and the antenna is too far off, they manually adjust it and measure again until it passes the accuracy check. This ensures the average is perfect, but it means the individual dishes end up scattered around that average, rather than all landing in the exact same tiny spot. The team admits that their current manual process for installing the antennas needs some engineering tweaks to make every single dish land in the same perfect spot every time.

The Final Verdict

So, what's the bottom line? The HIRAX team has successfully built a set of dishes that are mostly up to snuff. The surfaces are smooth enough, and the antennas are generally in the right place. The main takeaway is that the team has identified a small but important difference in the dishes made from the third mold and has figured out how to handle it. They also know that their manual installation process for the antennas needs to be tightened up to reduce the scatter.

These measurements aren't just a quality check; they are the first step in building a computer model of how the telescope will actually see the sky. By feeding these real-world measurements into their simulations, the scientists can predict exactly how the telescope's "beam" will look and correct for any errors before they even start their big survey. This careful, playful detective work ensures that when HIRAX finally listens for the hydrogen whispers, it won't be fooled by the stadium noise, allowing us to finally map the dark energy that is shaping our universe.

Technical Summary: Dish Assembly Precision for HIRAX

Problem Statement
The Hydrogen Intensity and Real-time Analysis eXperiment (HIRAX) is a radio interferometer array designed to map the large-scale structure of the universe by observing the 21 cm neutral hydrogen (HI) line via intensity mapping (IM) across the redshift range of 0.78–2.55. A critical challenge in extracting the faint cosmological 21 cm signal is the presence of astrophysical foregrounds that are up to six orders of magnitude brighter. Instrumental systematic errors, specifically dish surface deviations and feed placement errors, can distort the primary beam, causing foreground power to leak into the clean cosmological window. To mitigate this, HIRAX utilizes a highly redundant array configuration, which relies on the assumption that telescope pairs with identical baselines observe identical sky signals. This assumption demands strict accuracy and precision in the mechanical assembly of the telescope's 6-meter parabolic fiberglass dishes and their prime focus feeds.

Methodology
This paper presents a metrological analysis of the first 28 HIRAX dishes manufactured for the Swartfontein site in South Africa. The study employs Digital Photogrammetric Analysis (DPA) using a DPA Professional system and AICON 3D Studio software to acquire data. This method was selected for its ability to provide rapid, repeatable measurements with consistent point density, offering a significant speed advantage over laser tracking.

The analysis covers two primary areas:

Dish Surface Metrology: Measurements were taken both in the factory (during fabrication) and on-site (after installation on mounts). The data was used to fit a best-fit paraboloid to each dish to determine focal lengths and surface deviations.
Feed Placement Metrology: A dedicated verification jig was used to measure the feed's position relative to the dish's best-fit paraboloid. Parameters measured included translational offsets ($dx, dy, dz$), feed inclination ( $\delta$ ), and N-S alignment ( $\gamma$ ).

The study evaluates performance against the HIRAX Telescope Mechanical Assembly Requirements Document, defining accuracy as the absolute difference between the mean of measured values and the ideal value, and precision as the root-mean-square (RMS) error from the mean. Due to observed systematic differences in focal lengths based on manufacturing molds, the data was analyzed using two approaches:

Approach 1: Treating all 28 dishes as a single ensemble.
Approach 2: Segregating dishes into two groups based on the manufacturing mold (Molds 1 & 2 vs. Mold 3).

Key Results

Dish Surface Accuracy and Precision:
- Focal Length: The mean best-fit focal length for the first 28 dishes was $1261.6 \pm 1.5$ mm. A significant systematic shift was identified: dishes from Mold 3 had a mean focal length approximately 3 mm longer than those from Molds 1 and 2, attributed to the degradation of the master plug over time.
- Surface Accuracy: Under Approach 1 (single ensemble), the maximum absolute residual of the mean best-fit paraboloid was 2.3 mm, satisfying the 5.0 mm accuracy requirement. Under Approach 2, the accuracy for Molds 1 & 2 was 1.1 mm, while Mold 3 showed 5.4 mm.
- Surface Precision: Under Approach 1, only 36% of dishes met the 1.0 mm polar RMS precision requirement. However, under Approach 2 (separating by mold), 96% of dishes met the precision threshold, with only one dish failing.
- On-Site Measurements: On-site measurements of the first 19 dishes revealed a mean focal length of $1261.1 \pm 1.5$ mm, consistent with factory data. However, initial on-site tests indicated quadrupole-like surface deviations induced by mechanical stress from the mounts. Consequently, on-site polar RMS analysis was deferred pending the mitigation of these stress issues.
Feed Placement Accuracy and Precision:
- Accuracy: All feed parameters ( $dx, dy, dz, \delta, \gamma$ ) met the required accuracy thresholds.
- Precision: Only the feed inclination angle ( $\delta$ ) met the target precision of 2.5 arcmin (achieved 1.6 arcmin). The translational offsets ($dx, dy$) and N-S alignment ( $\gamma$ ) exhibited significant scatter, failing the precision targets (e.g., $dx$ precision was 1.2 mm vs. a 0.5 mm target; $\gamma$ precision was 3.8 arcmin vs. a 2.5 arcmin target).
- Cause of Scatter: The paper attributes the scatter in alignment parameters to the production process, where photogrammetry is used as an active feedback mechanism. If a dish fails accuracy checks, it is manually adjusted and remeasured. This iterative process ensures accuracy but introduces variability that impacts the statistical precision of the final dataset.
- Height ($dz$): The vertical placement was found to be repeatable, though outliers in Mold 3 dishes were noted due to a shift in the target definition from relative focal length to absolute z-height to ensure manufacturing repeatability.

Significance and Claims
The paper establishes a baseline for modeling primary beam effects, which is essential for characterizing and mitigating systematic errors in the HI IM signal for the full HIRAX array. The authors claim that while surface deviations are predominantly within requirements, the feed placement precision currently falls short of targets, though this is expected to improve with a larger sample size and refined protocols.

The study concludes that the choice of analytical framework (Approach 1 vs. Approach 2) has implications for the array's calibration strategy. If Approach 2 is adopted, redundant calibration cannot treat the array as a single uniform population but must accommodate two distinct sub-ensembles with different primary beam responses. The authors state that propagating these measured physical deviations into simulated primary beams is the necessary next step to determine which calibration approach provides the most robust suppression of systematics-induced foreground contamination, ultimately enabling precision cosmological constraints.

More like this