🔭 astrophysics

CHIME-o-Grav: Wideband Timing of Four Millisecond Pulsars from the NANOGrav 15-yr dataset

This paper presents the integration of high-cadence CHIME telescope data with NANOGrav observations to perform wideband timing on four millisecond pulsars, resulting in improved timing precision and updated measurements of relativistic parameters, including a precise mass determination for the companion of PSR J2302+4442 via Shapiro delay.

Original authors: Gabriella Agazie, David L. Kaplan, Abhimanyu Susobhanan, Ingrid H. Stairs, Deborah C. Good, Bradley W. Meyers, Emmanuel Fonseca, Timothy T. Pennucci, Akash Anumarlapudi, Anne M. Archibald, Zaven Arzou

Published 2026-03-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Gabriella Agazie, David L. Kaplan, Abhimanyu Susobhanan, Ingrid H. Stairs, Deborah C. Good, Bradley W. Meyers, Emmanuel Fonseca, Timothy T. Pennucci, Akash Anumarlapudi, Anne M. Archibald, Zaven Arzoumanian, Paul T. Baker, Paul R. Brook, Alyssa Cassity, H. Thankful Cromartie, Kathryn Crowter, Megan E. DeCesar, Paul B. Demorest, Timothy Dolch, Fengqiu Adam Dong, Elizabeth C. Ferrara, William Fiore, Gabriel E. Freedman, Nate Garver-Daniels, Peter A. Gentile, Joseph Glaser, Jeffrey S. Hazboun, Ross J. Jennings, Megan L. Jones, Matthew Kerr, Michael T. Lam, Duncan R. Lorimer, Jing Luo, Ryan S. Lynch, Alexander McEwen, James W. McKee, Maura A. McLaughlin, Natasha McMann, Cherry Ng, David J. Nice, Benetge B. P. Perera, Nihan S. Pol, Henri A. Radovan, Scott M. Ransom, Paul S. Ray, Alexander Saffer, Ann Schmiedekamp, Carl Schmiedekamp, Brent J. Shapiro-Albert, Kevin Stovall, Joseph K. Swiggum, Mercedes S. Thompson, Haley M. Wahl

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 is a giant, cosmic clock tower. Scattered throughout this tower are the most precise timekeepers imaginable: pulsars. These are dead stars (neutron stars) that spin incredibly fast, beaming lighthouse-like flashes of radio waves toward Earth. Because they spin so steadily, astronomers use them to test the laws of physics, hunt for gravitational waves (ripples in space-time), and even map the invisible stuff between stars.

However, these cosmic clocks aren't perfect. As their signals travel through space, they get "jittery" due to two main things:

The Interstellar Medium: A fog of gas and dust that slows down the radio waves, making them arrive a tiny bit late.
The Solar Wind: A stream of charged particles from our own Sun that acts like a temporary traffic jam, further delaying the signal.

The Problem: Missing the "Low Notes"

For years, the NANOGrav project (a team of astronomers in North America) has been listening to these pulsars using giant radio dishes like the Green Bank Telescope (GBT). They listen mostly at high frequencies (like the high notes on a piano). High notes are great for timing, but they don't tell you much about the "fog" (the gas) slowing them down.

To get a perfect picture, you need to listen to the low notes too. Low-frequency radio waves are slowed down much more by the gas, which actually helps astronomers measure exactly how much gas is in the way. But, until recently, we didn't have a telescope that could listen to these low notes every single day with high precision.

The Solution: The CHIME Telescope

Enter CHIME (the Canadian Hydrogen Intensity Mapping Experiment). Think of CHIME as a massive, stationary radio telescope that looks like a giant half-pipe. It doesn't move; instead, it watches the sky drift by as the Earth rotates. Since 2019, it has been listening to these pulsars every single day at low frequencies.

This paper is about what happens when you combine the "high notes" from NANOGrav with the "daily low notes" from CHIME.

The "Wideband" Magic

Usually, astronomers have to split their data into many small frequency slices to figure out the timing. It's like trying to solve a puzzle by looking at one tiny piece at a time.

The authors developed a new method called Wideband Timing. Imagine instead of looking at puzzle pieces one by one, you have a smart template of the whole puzzle. You can look at the entire range of frequencies at once and instantly figure out:

Exactly when the pulse arrived.
Exactly how much gas was in the way.

This is like having a super-smart translator that can hear a whole sentence and instantly know both the meaning and the accent, rather than guessing word by word.

What They Found (The Four Stars)

The team applied this new method to four specific pulsars, mixing the daily CHIME data with the long-term NANOGrav data. Here's what they discovered:

The "Heavy" Star (PSR J2302+4442):
This pulsar has a companion star orbiting it. By using the new data, they could measure the "Shapiro delay"—a fancy term for how the companion's gravity bends space and slows down the signal.
- The Result: They finally got a very precise weight for both stars. The companion is a small white dwarf (about 1/3 the mass of our Sun), and the pulsar is a heavy neutron star (almost twice the mass of our Sun). It's like finally getting a precise reading on a bathroom scale after years of guessing.
The "Orbiting" Stars (PSR J1012+5307 & PSR J2145−0750):
These stars are in binary systems. The team measured how their orbits are changing over time.
- The Result: They confirmed that these orbits are shrinking exactly as Einstein's theory of General Relativity predicts (due to the emission of gravitational waves). The new data made these measurements much sharper, like switching from a blurry photo to a 4K image.
The "Solo" Star (PSR J0645+5158):
This one isn't in a binary system. The team used the new data to calculate exactly how fast it's slowing down its spin.
- The Result: They stripped away all the external factors (like the galaxy's gravity) to find the star's "true" internal slowdown rate.

The Solar Wind Surprise

One of the coolest side discoveries involved the Solar Wind. Because CHIME listens every day, the team could see how the Sun's wind changes the pulsar signals in real-time.

The Analogy: Imagine trying to time a runner while a wind gust hits them every year. Previous models assumed the wind was a steady, predictable breeze. The new data showed the wind is actually chaotic and lumpy. The CHIME data revealed that the "wind" hitting these pulsars is much more complex than we thought, creating weird bumps in the timing data that old models couldn't explain.

Why Does This Matter?

Better Gravity Tests: By knowing the exact timing and the exact amount of gas in the way, we can test Einstein's theories of gravity with unprecedented precision.
Gravitational Waves: To find the "hum" of gravitational waves from supermassive black holes, we need to remove all the noise. The new method removes the "gas noise" much better than before.
Data Volume: CHIME produces so much data (daily vs. monthly) that it's like switching from a dial-up internet connection to fiber optics. It's a massive upgrade for the future of astronomy.

In a Nutshell

This paper is a success story of teamwork between telescopes. By combining the high-precision, long-term listening of NANOGrav with the daily, low-frequency "ears" of CHIME, and using a clever new math trick (Wideband Timing), the team has turned four cosmic clocks into even more precise instruments. They've weighed stars more accurately, confirmed Einstein's predictions, and learned that the solar wind is much more turbulent than we ever imagined.

1. Problem Statement

Pulsar Timing Arrays (PTAs) like NANOGrav rely on long-term, high-precision timing of Millisecond Pulsars (MSPs) to detect nanohertz-frequency gravitational waves. A major challenge in PTA analysis is handling time-varying Dispersion Measure (DM) caused by the interstellar medium (ISM) and frequency-dependent pulse profile evolution. Traditional narrowband timing splits observations into sub-bands, requiring multiple Time-of-Arrival (TOA) measurements per epoch and introducing arbitrary frequency-dependent parameters to model profile evolution.

Furthermore, the Canadian Hydrogen Intensity Mapping Experiment (CHIME) telescope has been observing NANOGrav MSPs at a nearly daily cadence (compared to the monthly cadence of other telescopes) since 2019. However, integrating this high-cadence, low-frequency (400–800 MHz) data with existing high-frequency datasets (like the NANOGrav 15-yr dataset, NG15) presents challenges:

CHIME data has lower gain and higher TOA uncertainties due to lower observing frequencies.
The sheer volume of TOAs generated by daily observations creates data management issues.
Standard narrowband methods are inefficient for such high-cadence, wideband data.

The paper addresses how to effectively integrate CHIME data into the NG15 dataset using Wideband (WB) timing to improve timing precision, characterize DM variations, and refine relativistic parameters for binary pulsars.

2. Methodology

Data Sources and Reduction

Datasets: The study combines the NANOGrav 15-yr (NG15) dataset (primarily from the Green Bank Telescope, GBT) with CHIME data (April 2019 – March 2023).
Targets: Four MSPs were analyzed:
- PSR J0645+5158 (Isolated)
- PSR J1012+5307 (Binary)
- PSR J2145−0750 (Binary)
- PSR J2302+4442 (Binary)
Pipeline:
- NG15 Data: Processed using the nanopipe pipeline.
- CHIME Data: Processed using a newly developed pipeline called chimera (Susobhanan 2026). This pipeline performs RFI excision, flux/polarization calibration (noting that polarization calibration for CHIME is still under development and was omitted for this analysis), and frequency/time scrunching.
- Wideband TOA Calculation: Both datasets utilized the PulsePortraiture software suite. Instead of splitting bands, a 2-D frequency-dependent "template portrait" (constructed via Principal Component Analysis and spline interpolation) is used to simultaneously solve for a single TOA and a single DM for each observation.

Timing Analysis

Software: Timing was performed using PINT.
Binary Modeling:
- PSR J1012+5307: Modeled with the ELL1 model (low eccentricity).
- PSR J2145−0750 & PSR J2302+4442: Modeled with the ELL1H model to account for significant Shapiro delay detections.
DM Modeling: Used the piece-wise constant DMX model. The authors tested "constant" vs. "hybrid" binning strategies (using shorter bins for high-cadence CHIME data) and selected the best model using the Akaike Information Criterion (AIC).
Noise Modeling: Used the enterprise Bayesian suite to characterize white noise (EFAC, EQUAD) and red noise (Gaussian process).
Solar Wind: Attempted to model solar wind DM variations for PSR J2145−0750, finding that standard radial models were insufficient; denser DMX binning was used as a proxy.

3. Key Contributions

Development of the chimera Pipeline: A specialized data reduction pipeline for generating Wideband TOAs from CHIME/Pulsar backend data, enabling its integration into PTA datasets.
First Large-Scale WB Integration: This is one of the first major efforts to combine CHIME data with a PTA dataset (NG15) using the Wideband paradigm, tripling the TOA volume for some sources.
Improved Relativistic Measurements: Significant improvements in Post-Keplerian (PK) parameters, specifically orbital decay ( $\dot{P}_b$ ) and the rate of change of the projected semi-major axis ( $\dot{x}$ ), driven by the extended timing baseline.
Precise Mass and Inclination Constraints:
- PSR J2302+4442: Provided a strong detection of Shapiro delay, yielding precise measurements of companion mass, pulsar mass, and orbital inclination.
- PSR J1012+5307 & PSR J2145−0750: Combined timing data with Very Long Baseline Interferometry (VLBI) astrometry to constrain system inclinations and masses.

4. Key Results

General Timing Statistics

TOA Volume: CHIME data nearly tripled the total TOA count for the analyzed sources (e.g., ~1,000 CHIME TOAs vs. ~200–600 NG15 TOAs per source).
Precision:
- TOA Uncertainty: CHIME WB TOAs generally have larger uncertainties than GBT TOAs due to lower frequency ( $\nu^{-1/2}$ scaling) and lower gain.
- DM Precision: Conversely, CHIME WB DMs are significantly more precise than GBT DMs. The large bandwidth (400 MHz) and low-frequency coverage provide a much larger lever arm for measuring DM delays ( $\nu^{-2}$ scaling).

Specific Source Results

PSR J0645+5158 (Isolated):
- Updated parallax distance: $1.2(2)$ kpc.
- Intrinsic spin-down ( $\dot{P}_{int}$ ) calculated as $4.5(12) \times 10^{-21}$ s s $^{-1}$ after correcting for kinematic and galactic acceleration effects.
PSR J1012+5307 (Binary):
- Orbital Decay ( $\dot{P}_b$ ): Precision improved by a factor of 5 compared to NG15. The measured $\dot{P}_b$ is consistent with General Relativity (GR) predictions after correcting for Shklovskii and galactic acceleration effects using VLBI astrometry.
- $\dot{x}$ (Projected semi-major axis derivative): Precision improved by a factor of 3. The measurement is dominated by proper motion ( $\dot{x}_{PM}$ ), providing constraints on the orbital inclination ( $i \leq 52^\circ \pm 4^\circ$ in the $0-90^\circ$ quadrant).
PSR J2145−0750 (Binary):
- $\dot{x}$ : Improved precision by a factor of 4, but the value shifted significantly (12.5- $\sigma$ from NG15). The authors attribute this to the inability to model red noise (likely due to solar wind variability) in the combined dataset.
- Shapiro Delay: Detected the $h_3$ harmonic. Combined with VLBI data, the results are consistent with previous mass/inclination measurements.
PSR J2302+4442 (Binary):
- Shapiro Delay: Achieved a strong detection allowing for direct mass measurements.
- Companion Mass: $0.35^{+0.05}_{-0.04} M_\odot$
- Pulsar Mass: $1.8^{+0.3}_{-0.3} M_\odot$
- Inclination: $80^\circ^{+1}_{-2}$
- These results represent a dramatic improvement (orders of magnitude) in pulsar mass uncertainty compared to previous NG15 results.

Solar Wind Modeling

For PSR J2145−0750, the high-cadence CHIME data revealed asymmetric solar wind peaks in the DM time series that standard radial models could not capture. The authors found that increasing DMX bin density was the most robust method to handle these variations for this specific source.

5. Significance and Future Implications

PTA Sensitivity: The inclusion of CHIME data significantly extends the timing baseline and increases the number of observations, which is critical for improving the sensitivity of PTAs to gravitational waves.
Data Volume Management: The study demonstrates that Wideband timing is essential for handling the "data explosion" from high-cadence telescopes like CHIME. It reduces the number of TOAs per epoch (from dozens in narrowband to one in wideband) while maintaining or improving DM precision.
Interstellar Medium (ISM) Studies: The high-precision, low-frequency DM measurements from CHIME provide a unique laboratory for studying ISM turbulence and solar wind effects, which are major noise sources in PTA analyses.
Future Work: The authors note that future PTA datasets (e.g., NANOGrav 20-yr, IPTA 3rd release) will heavily rely on CHIME data. Future work will focus on developing better RFI excision algorithms for CHIME and more sophisticated solar wind models to further reduce timing residuals.

In conclusion, this paper successfully validates the integration of high-cadence, low-frequency CHIME data into PTA analyses using Wideband timing, yielding superior constraints on binary system parameters and demonstrating the critical role of CHIME in the future of gravitational wave astronomy.