Reis Tyson

01

Constraining
Dark Energy
with BAO

From a cosmological standard ruler to a two-parameter dark-energy constraint using χ² optimisation

Δχ² landscape

Hover parameter space to inspect the fit.

wₚ-0.9592wₐ-0.1946Δχ²0.000
-1.00-0.95-0.90-0.350.000.35wₚwₐ
● Best fit× ΛCDM— 1σ contour– – 2σ contour

-0.9592

Best wₚ

-0.1946

Best wₐ

15.12

Minimum χ²

0.94

Reduced χ²

These constraints are produced from simplified mock datapoints and demonstrate the inference procedure only. They are not measurements of the real Universe.

Baryon Acoustic Oscillations

A fixed physical scale can become a cosmological measuring tool.

Baryon acoustic oscillations preserve a characteristic scale in the large-scale distribution of matter. Unlike a standard candle, whose known luminosity is compared with its observed brightness, BAO acts as a standard ruler: a known physical scale whose apparent size depends on cosmological distance.

If the expansion history changes, the inferred distance to a given redshift changes. The same 150 Mpc ruler then subtends a different angle, allowing cosmological parameters to be tested against the observed BAO signal.

Core relation

θBAO = lBAOd(z)

lBAO = 150 Mpc

d(z) is the cosmology-dependent comoving distance.

θBAO is the corresponding angular scale.

Standard Ruler

Connect redshift to distance, then distance to angle.

For each redshift, the expansion history determines the comoving distance. Dividing the fixed BAO scale by that distance gives the predicted angular scale.

Move through the mock redshift sample to see how the same physical ruler is viewed at different cosmological distances.

Standard ruler

ΛCDM reference

150 MpcθOBSERVERCOSMOLOGICAL DISTANCE

Redshift

1.2

z

z = 0.3z = 2.0
Comoving distance3872.1 Mpc
BAO scale150 Mpc
Predicted angle0.038738
Mock Δθ+2.000e-5

The physical BAO ruler remains fixed at 150 Mpc. Increasing redshift moves the observer farther from that ruler in this illustrative geometry, reducing the angle it subtends.

Diagram geometry is illustrative; the distance and angular values are calculated from the cosmological model.

Mock Measurements

Start with the measurements the fit is actually asked to explain.

The analysis uses 18 simple mock BAO measurements between z = 0.3 and z = 2.0.

Important

These are simplified mock datapoints used to demonstrate the cosmological inference method. They are not measurements of the real Universe and should not be interpreted as a real Euclid cosmological constraint.

What is measured?

Mock BAO angular residual relative to the Lambda-CDM reference cosmology.

z

Redshift

Position of the mock observation in cosmic history.

Δθ

Angular residual

Difference from the reference ΛCDM BAO prediction.

σθ

Uncertainty

A fixed 5.0e-5 uncertainty is used in this demonstration.

Complete input data

All values used by the demonstration.

Simulated data
PointRedshift zMock Δθσθ
010.3+3.420e-45.00e-5
020.4+2.490e-45.00e-5
030.5+2.280e-45.00e-5
040.6+1.410e-45.00e-5
050.7+1.270e-45.00e-5
060.8+6.200e-55.00e-5
070.9+1.180e-45.00e-5
081.0+9.900e-55.00e-5
091.1+2.800e-55.00e-5
101.2+2.000e-55.00e-5
111.3+1.500e-55.00e-5
121.4+4.700e-55.00e-5
131.5+4.900e-55.00e-5
141.6-1.800e-55.00e-5
151.7+2.300e-55.00e-5
161.8+1.000e-65.00e-5
171.9-4.000e-55.00e-5
182.0-1.770e-45.00e-5

Infer Cosmology

Change the expansion history and the BAO prediction moves.

The model allows dark energy to evolve with scale factor using

w(a) = wp + wa(1 − a)

Each pair (wₚ, wₐ) produces a different distance-redshift relation and therefore a different set of BAO angular residuals.

01

Choose cosmology

Select a trial pair of dark-energy parameters within wₚ = -1.00 to -0.90 and wₐ = -0.35 to 0.35.

02

Infer distance

Numerically integrate the expansion history to obtain d(z) for every mock redshift.

03

Predict BAO

Convert each cosmological distance into an angular BAO prediction and subtract the ΛCDM reference.

04

Score the model

Compare the predicted residuals with the mock observations using χ².

From model to prediction

The best-fit cosmology traces the mock BAO residuals across redshift.

Points show the simplified mock observations. The line shows the angular residual predicted by the optimised cosmological model.

Preliminary Results

The minimum is only one point. The surrounding landscape tells us what is constrained.

The optimisation identifies the lowest χ², while a parameter grid maps how rapidly the fit deteriorates away from that solution.

Joint confidence regions for two fitted parameters are defined by Δχ² ≤ 2.30 at 1σ and Δχ² ≤ 6.18 at 2σ.

Δχ² landscape

Hover parameter space to inspect the fit.

wₚ-0.9592wₐ-0.1946Δχ²0.000
-1.00-0.95-0.90-0.350.000.35wₚwₐ
● Best fit× ΛCDM— 1σ contour– – 2σ contour

Inferred Cosmology

Translate the χ² minimum into an inferred dark-energy model.

For this mock dataset, the numerical optimiser identifies the parameter pair that most closely reproduces the supplied BAO residuals.

Interpretation limit

These values are a demonstration of parameter inference from simplified mock measurements. They must not be interpreted as a measurement of the actual dark-energy equation of state.

Best-fit wₚ

-0.9592

Best-fit wₐ

-0.1946

Minimum χ²

15.116

16 degrees of freedom

Reduced χ²

0.945

Projected parameter ranges

wₚ -0.9660 -0.9527wₐ -0.2380 -0.1540
wₚ -0.9700 -0.9480wₐ -0.2660 -0.1260

ΛCDM reference

The ΛCDM point (wₚ = −1, wₐ = 0) has Δχ² ≈ 118.62 on the sampled grid and lies outside the plotted 2σ region.

Implementation

Small enough to expose the complete numerical workflow.

The analysis is intentionally compact. The published Python implementation performs the cosmological distance integration, χ² minimisation, confidence mapping and website-data export.

A compact NumPy archive preserves the numerical output, while the JSON export supplies the values displayed on this page.

This study is a methodological demonstration using simplified mock BAO measurements. It demonstrates the path from standard-ruler observations to cosmological parameter inference, not a contemporary Euclid cosmological result.

BAOχ² optimisationDark energyMock dataNumerical inference