Mobirise

Multiprobe Cosmology 

We combine different cosmological probes and datasets to explore fundamental physics in the Universe. 

Learn More
Mobirise

Kinematic Lensing

KL combines imaging and spectroscopic data into a new type of shear inference. It is a promising avenue for cosmological measurements.

Learn More
Mobirise

Inflation 

The NASA SPHEREx satellite mission will constrain the physics of inflation, a rapid expansion of space, in the early Universe.

Learn More
Mobirise

Machine Learning

We use ML to accelerate complex calculations of cosmological models and to learn features in large datasets.

Learn More

Multi-Probe Cosmology - Exploring the Universe with cosmological surveys


Our lab is interested in exploring fundamental physics questions about our Universe, dark energy, dark matter, gravity, neutrino mass, and inflation:

1) What is the underlying physical mechanism driving the accelerated expansion of the Universe?
2) What is the mass and number of neutrino species?
3) What is the nature of dark matter and the connection of dark matter overdensities and galaxy formation?
4) Is General Relativity a complete description of our theory of gravity or does it need to be modified as a function of the environment or the distances involved?
5) Can we measure the physics of inflation, i.e. of the very first second when the Universe began?

We explore these questions through a combination of theoretical modeling and statistical data analysis.

Selected papers on Dark Energy Survey data analysis led by or with major contributions from the ACL

Selected papers on Rubin's LSST led by or with major contributions from the ACL

Selected papers on the Roman Space Telescope led by or with major contributions from the ACL


Kinematic Lensing (KL)

Weak lensing measurements are extremely challenging: the shear effect is degenerate with the unobservable intrinsic galaxy shape (with shape dispersion σ_ε~0.25). This corresponds to a S/N ~ 0.01 per galaxy and makes traditional weak lensing a statistical measurement that requires averaging over large numbers of background galaxies.
Kinematic lensing breaks the shape-shear degeneracy by combining imaging and spectroscopic data into a new type of lensing inference.
Spiral galaxies exhibit disk rotation patterns with a circular velocity that can be inferred from the easily observable galaxy luminosity (Tully-Fisher relation, Huff et al. 2013). Spectroscopic measurements of the rotation velocity of a disk galaxy enable us to infer the unobservable intrinsic galaxy shape in weak lensing measurements and measure shear on a per-galaxy basis.

Mobirise

Kinematic Lensing (KL) - Basic Concepts

The basic idea for the kinematic lensing is depicted on the right. In an image, an inclined rotating circular disk has elliptical isophotes. When the image of this galaxy is sheared, the isophotes remain elliptical (in the weak shear limit, |g| << 1) with a new axis ratio and position angle. New photometric axes are inferred from this ellipse in the sheared image, and information about the original axes is lost. The case is different however, with galaxy kinematics measured via spectroscopy. The unsheared circular disk has kinematic axes that are perpendicular to one another and are aligned with the unsheared photometric axes. This cross shape becomes skewed when the velocity map is sheared; the kinematic axes are no longer perpendicular and they are misaligned with the photometric axes inferred from the sheared isophotal ellipse in the imaging data.

The fact that the velocity map transforms differently than the photometric image when shear is applied can be used to combine imaging and spectroscopic information to tightly constrain the intrinsic ellipticity of the source galaxies, thereby removing the largest statistical uncertainty in cosmic shear.

Mobirise

Kinematic Lensing - Roman Space Telescope

In Xu et al 2023 we explore the KL concept in the context of the Roman Space Telescope and find that it would significantly boost the cosmological information compared to the Roman traditional WL measurement. We note that KL and WL are complementary since the underlying galaxy samples are fundamentally different.


The figure to the right shows the improvement of KL (blue contours) with Roman ST over WL with Roman ST (black solid), which translates into an improvement of a factor of 3.65 in the relevant figure-of-merit, even though the KL galaxy sample is > a factor of 10 smaller.  

Mobirise

Kinematic Lensing - Pilot Measurement

To demonstrate the KL technique, the ACL team has developed an analysis pipeline for long-slit spectroscopic data. Per galaxy we fit 11 model parameters: maximum circular velocity, galaxy systemic velocity, galaxy inclination, image half-light radius, spectrum half-light radius, velocity scale radius, central brightness, and disk scale height. Pranjal RS et al. 2023 demonstrated unbiased shear inference on simulated data and provide our most realistic shape noise estimates for KL (c.f. figure on the right).
We are currently applying this KL inference pipeline to Keck DEIMOS spectra of source galaxies in the Abel 2261 field for a pilot measurement.

Selected papers on KL led by or with major contributions from the ACL

Mobirise

Inflation - SPHEREx mission

SPHEREx (Spectro-Photometer for the History of the Universe, Epoch of Reionization, and Ices Explorer) is a selected NASA explorer mission that will observe the entire sky with near-spectroscopic resolution. By measuring a unique signal in the clustering of galaxies on large scales, SPHEREx will be able to constrain whether the initial density field that seeds our Universe has signatures of non-Gaussianity. These signatures allow us to constrain different physics scenarios of "inflation" (see Fig to the right), a rapid expansion phase in the early Universe.

The mission is led by Caltech/JPL (SPHEREx mission website), our group is building the modeling and inference code for the galaxy power spectrum analysis. We are also working on quantifying synergies of SPHEREx and other cosmological missions (in particular LSST) with the goal to constrain Dark Energy.

Selected papers on SPHEREx science led by or with major contributions from the ACL

Mobirise

Machine Learning

The challenge of balancing accuracy and speed in modeling multi-probe data vectors has been a limiting factor already for the current generation of cosmological surveys. This problem will increase significantly given the decrease in statistical uncertainty for future datasets and the corresponding increase in model complexity.  

Emulating the results of expensive numerical simulations and subsequent complex calculations that underpin the computation of observables as a function of the parameter space has been identified as the most promising path forward. We are developing neural network architectures that are precise, fast, and computationally inexpensive in performing these computations and that will enable the complex analyses using Roman, Rubin, SPHEREx, and DESI data. 

Selected papers using ML with major contributions from the ACL

‌

Landing Page Software