🔭 astrophysics

Closed-Form Statistical Relations Between Projected Separation, Semimajor Axis, Companion Mass, and Host Acceleration

This paper derives and verifies closed-form, analytic probability density functions that statistically relate a companion's mass, projected separation, and host acceleration without dependence on specific Keplerian orbital elements, offering computationally efficient tools for exoplanet characterization.

Original authors: Timothy D Brandt

Published 2026-04-17

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Timothy D Brandt

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 you are standing in a dark room, watching a faint, invisible dancer (a star) and a tiny, invisible partner (a companion planet or brown dwarf) spinning around each other. You can't see the full dance because you've only been watching for a few seconds, but you can see two things:

The Tug: The star is wobbling or speeding up slightly in a specific direction (this is the "acceleration" or "trend").
The Distance: You can see the companion at a certain distance from the star on the flat screen of your vision (this is the "projected separation").

The big question astronomers have always asked is: "How heavy is the invisible partner?"

For a long time, figuring this out was like trying to guess the weight of a mystery box by shaking it, but you didn't know how hard you were shaking it, how far the box was from the center, or what angle it was spinning at. Scientists had to run massive, slow computer simulations (like rolling dice millions of times) to get a rough answer.

The "Magic Formula" Breakthrough

In this paper, the author, Timothy Brandt, says: "Stop rolling dice. I found a direct math formula."

He derived a set of "closed-form" equations. In plain English, this means he found a simple, direct recipe to calculate the probability of the companion's mass without needing to simulate thousands of random orbits.

Here is how he did it, using some creative analogies:

1. The "Random Orientation" Trick

Imagine throwing a handful of darts at a giant, invisible sphere. Most darts will land near the "equator" (the middle), and fewer will land near the "poles" (the top and bottom).

The Problem: We don't know if the star's companion is orbiting like a dart near the pole (head-on) or near the equator (side-on).
The Solution: Brandt realized that if you assume the orbit is randomly oriented in space (like those darts), you can ignore all the complicated details of the orbit (like how stretched out it is or where it started). You only need to know the angle between our line of sight and the companion.
The Result: He turned this geometric probability into a simple math function. It's like realizing that no matter how the dancer spins, the average wobble you see follows a predictable pattern based purely on geometry.

2. The "Speed vs. Distance" Balance

The paper connects three things:

The Wobble (Acceleration): How fast the star is being pulled.
The Gap (Projected Separation): How far apart they look to us.
The Weight (Mass): How heavy the companion is.

Brandt's formula says: If you see a certain wobble at a certain distance, here is the exact statistical chance that the companion is a giant planet, a brown dwarf, or a small star.

He showed that his new, simple formula gives the exact same answer as the old, complicated computer simulations, but it's much faster. It's like replacing a 10-hour cooking process with a microwave recipe that tastes exactly the same.

3. The "Shadow" Analogy (The Ratio Problem)

The second half of the paper tackles a different puzzle: "How far apart are they really compared to how far apart they look?"

Imagine a lighthouse beam. If the beam points straight at you, it looks like a dot. If it points sideways, it looks like a long line. The "projected separation" is just the shadow the orbit casts on our sky.

The author derived a formula to calculate the probability of the "real" distance versus the "shadow" distance.
He used a special type of math called Elliptic Integrals (think of these as "super-advanced rulers" for curved shapes) to solve this.
Why it matters: This helps astronomers know if a companion they see is actually close to the star or if it's just a trick of perspective.

Why Should You Care?

Speed: In the age of big data (like the Gaia space telescope sending back millions of star positions), we need to calculate these probabilities instantly. Old methods were too slow; Brandt's formulas are instant.
Precision: Because the formulas are "analytic" (pure math), they are perfect for advanced computer algorithms (like Hamiltonian Monte Carlo) that need to take derivatives (measure how fast things change) to find the most likely answer.
Simplicity: He proved that you don't need to know the complex history of an orbit to guess the mass of a companion. You just need the geometry of the moment.

The Bottom Line

Timothy Brandt took a problem that required a supercomputer to simulate millions of random orbits and turned it into a simple, elegant math equation. It's the difference between guessing the weight of a fish by throwing it in the air and catching it a million times, versus just measuring the water displacement with a ruler.

He even made a "cookbook" (a Jupyter notebook) available online so anyone can use these new, fast formulas to study the hidden companions of stars.

1. Problem Statement

Stars with widely separated companions often exhibit orbital periods longer than available observational baselines. Consequently, astronomers observe only a short segment of the orbit, appearing as an acceleration (or "trend") in the host star's motion.

Observables: Radial Velocity (RV) acceleration ( $a_{RV}$ ), Astrometric acceleration ( $a_{ast}$ ), and the projected separation ( $D\rho$ ) between the host and companion.
The Challenge: Determining the companion's mass ( $M_B$ ) and the ratio of projected separation to semimajor axis ( $\psi = D\rho/a$ ) from these observables is complex. Previous methods (e.g., Torres 1999, T99) relied on Monte Carlo simulations based on Keplerian orbital equations, which are computationally expensive and lack analytic derivatives.
Goal: To derive closed-form analytic expressions for the probability density functions (PDFs) of companion mass and separation ratios that are independent of specific Keplerian orbital elements, assuming random orbital orientations.

2. Methodology

The author derives analytic solutions by treating the separation vector between the star and companion as randomly oriented in space. This geometric assumption allows the decoupling of the problem from specific orbital elements (eccentricity, inclination, argument of periastron, etc.) for the mass derivation.

A. Mass Derivation (Section 2)

The author starts with Newton's Law of Gravity to relate mass, acceleration, and separation:

RV Case: $a_{RV} \propto M_B \cos\phi / r^2$
Astrometric Case: $a_{ast} \propto M_B \sin\phi / r^2$
Geometry: The projected separation is $D\rho = r \sin\phi$ , where $\phi$ is the angle between the line-of-sight and the separation vector.

By combining these, the mass $M_B$ is expressed as a function of the observables and a geometric factor $\Phi(\phi)$ .

Astrometric Factor: $\Phi_{ast} = 1/\sin^3\phi$ .
RV Factor: $\Phi_{RV} = 1/(\cos\phi \sin^2\phi)$ .

Assuming a random orientation (uniform distribution of $\cos\phi$ ), the author derives the Cumulative Distribution Functions (CDFs) and Probability Density Functions (PDFs) for $\Phi_{ast}$ and $\Phi_{RV}$ analytically.

Key Result: The PDFs are derived in closed form (Equations 8 and 17) without dependence on orbital elements, provided the orientation is random.

B. Equivalence to Keplerian Models (Sections 3 & 4)

To validate the analytic approach, the author:

Algebraic Proof: Demonstrates that the derived $\Phi_{RV}$ is mathematically equivalent to the complex Keplerian expression from T99 when averaged over a uniform distribution of orbital phases and arguments of periastron.
Empirical Verification: Generates $10^7$ Monte Carlo simulations with random orbital parameters (phase, inclination, eccentricity up to 0.8). The resulting histograms of the probability densities match the derived analytic curves perfectly (Figure 1).

C. Projected Separation to Semimajor Axis Ratio (Section 5)

The paper addresses the distribution of $\psi = (D\rho)/a$ .

Step 1: Derives the PDF for the true separation ratio $y = r/a$ using Kepler's equation and the eccentric anomaly.
Step 2: Converts $y$ to the projected ratio $\psi = y \sin\phi$ using the probability distribution of a product.
Result: The resulting PDF involves complete elliptic integrals of the first kind ( $K$ ).
- For $\psi < 1-e$ : Solution involves $K(\alpha)$ .
- For $1-e < \psi < 1+e$ : Solution involves $K(1/\alpha)$ .
Integration: While the solution is closed-form for a fixed eccentricity, integrating over an arbitrary eccentricity distribution ( $p(e)$ ) requires numerical quadrature (Equation 37).

3. Key Contributions

Analytic Mass PDFs: Provided the first closed-form expressions for the probability distribution of companion mass given RV or astrometric trends and projected separation. These expressions are significantly simpler than previous Monte Carlo-based approaches.
Derivative-Friendly: Because the results are analytic, they can be easily differentiated. This is crucial for modern statistical inference methods like Hamiltonian Monte Carlo (HMC), which require gradients for efficient sampling.
Separation Ratio PDFs: Derived closed-form solutions for the distribution of the projected separation to semimajor axis ratio ( $\psi$ ) at fixed eccentricity, expressed via elliptic integrals.
Validation: Rigorously proved that these simple geometric derivations are statistically equivalent to the full Keplerian orbital mechanics used in prior literature (T99).

4. Results

Mass Distributions: The analytic PDFs for mass (Equations 8 and 17) were shown to be indistinguishable from Monte Carlo results generated from T99's complex Keplerian equations.
Discrepancy Resolution: The author notes that T99's Figure 2 (inferred mass distributions) appeared different from the analytic results, specifically regarding a tail of inferred masses down to zero. The author confirms their analytic results are correct and that the T99 tail was likely an artifact of their specific implementation or sampling, not a physical necessity.
Separation Ratios: The paper provides 2D probability density maps of $\psi$ vs. eccentricity (Figure 2) and 1D distributions for various eccentricity priors (Figure 3). The distributions show an integrable singularity at $\psi = 1-e$ .

5. Significance

Computational Efficiency: Replacing Monte Carlo simulations with closed-form analytic expressions drastically reduces computational cost. This is vital for large-scale surveys (e.g., Gaia, direct imaging) where millions of systems must be analyzed.
Statistical Inference: The ability to compute derivatives of the likelihood function enables the use of gradient-based samplers (HMC), leading to faster convergence and more robust parameter estimation for exoplanet and brown dwarf companions.
General Applicability: The framework applies to any system where only a short orbital arc is observed, allowing for immediate statistical constraints on companion masses without needing to solve the full orbital solution.
Reproducibility: The author provides a Jupyter notebook with all code and figures, facilitating immediate adoption by the community.

In summary, this paper transforms a computationally intensive, simulation-heavy problem in exoplanet statistics into a set of elegant, analytic formulas, bridging the gap between geometric probability and Keplerian dynamics.