## Emmanuel Schaan

### eschaan@lbl.gov

I am a **Chamberlain postdoctoral fellow** at the **Lawrence Berkeley National Laboratory** since September 2017. I did my PhD at Princeton University with Professor David Spergel, and my undergraduate studies at Ecole Normale Supérieure de Paris.

See my **Publications** and **résumé**.

**Lower bias lower noise CMB lensing with shear-only reconstruction**

CMB foregrounds can be confused with lensing magnification but not with lensing shear. A shear-only CMB lensing estimator is thus immune to foregrounds. The signal-to-noise lost by discarding the magnification information is more than recovered by the ability to use higher temperature multipoles.

**Weak lensing of intensity maps: the case of the cosmic infrared background (CIB)**

Intensity mapping promises to efficiently sample the large-scale structure over a large fraction of the observable universe. However, the resulting intensity maps require complex modeling of the relationship between matter density and emissivity, and are affected by foregrounds/interlopers and missing modes. For these reasons, lens reconstruction from intensity maps would be useful. We showed how to measure lensing from any weakly non-Gaussian source field, and that current Planck CIB data is sufficient to detect this effect.

**First detection of the correlation between Lyman-alpha forest and CMB lensing**

Does the Lyman-alpha transmission trace the matter density in the way we typically assume? Can CMB help?** **We used CMB lensing as a tracer of the matter, to try to constrain the relation between Lyman-alpha transmission and matter density. We detected for the first time the bispectrum between the Lyman-alpha squared and the CMB lensing.

**Optical shear calibration with CMB lensing**

Can CMB help constrain galaxy lensing systematics?** **A lot of effort is invested in future lensing surveys (LSST, Euclid, WFIRST). Lensing systematics are challenging, e.g. shear calibration, photo-z estimation, intrinsic alignments, uncertainties in the matter power spectrum. Shear multiplicative bias can mimic a redshift dependence of the growth of structure. We made the most realistic forecast to date for CMB S4 lensing and LSST, and showed that CMB lensing can calibrate the shear biases for LSST. This is a self-calibration, purely empirical, and does not rely on any assumption about the galaxy population and morphologies, contrary to image simulations.

**kSZ measurement with ACTPol and BOSS**

How are the baryons distributed on the scale of a halo? Can CMB help?** **The matter power spectrum on small scales is uncertain, because of complex galaxy formation physics which changes the baryon density profile inside halos (“baryonic 1-halo term”). We detected the kSZ effect with a somewhat novel method, estimating velocities from the density field (as in BAO reconstruction). Our small signal-to-noise is already sufficient to show that the gas profile is more extended than the NFW profile.

**A stringent test of the effective field theory of large-scale structure**

Can we predict analytically the effect of non-linearities on the matter density field on small scales? We compared the EFT to N-body simulation at the level of the field instead of just the power spectrum. This allows for a comparison without sample variance. It is also a more stringent test, since power spectra are smooth functions, easy to match, and since two fields may have the same power spectrum without being identical. We showed that the EFT counterterms make the perturbative expansion convergent.

**Supersample covariance for n-point functions and cluster number counts**

Non-linear evolution makes the density field non-Gaussian and takes the information away from the power spectrum. What is the best way to extract information from this non-Gaussian density field? How can one recover some of this information, by combining power spectrum with higher n-point functions and halo counts? How to jointly analyze observables consistently, without double-counting the information? We used the halo model to compute the non-Gaussian covariances between matter/lensing n-point functions and halo counts, including the supersample covariance.

**Some fun stuff...**

Sailing around France on a 30ft boat

Photographing the Earth from a stratospheric balloon

Building a functional BB8 droid

(sort of) Flying a remote-controlled plane with the Peyton grad students

Hummingbird psychology... coming soon