Comparison of stats in no-baryons scenario (scale: ~10arcmin)
| Statistic | FoM |
|---|---|
| PS | 102 |
| Peaks | 227 |
| l1-norm | 353 |
\mathrm{FoM} \propto \dfrac{1}{\sqrt{ \det (\mathbf{C}_{\rm marg} )}}
Bias from baryons
θ~10arcmin/l=1024
θ~20arcmin/l=540
| Statistic | FoM |
|---|---|
| PS | 102 |
| Peaks | 219 |
| l1-norm | 288 |
| Statistic | FoM |
|---|---|
| PS | 53 |
| Peaks | 61 |
| l1-norm | 131 |
| Statistic | FoM(full)/FoM(cuts) |
|---|---|
| PS | 1.9 |
| Peaks | 3.6 |
| l1-norm | 2.2 |
Zu ̈rcher+ (2023)
Mask-dependence
Mask-dependence
(l1-norm)
Bias as function of survey area
BNT transform
- When we observe shear, contributions come from mass at different redshifts.
- BNT Transform: method to “null” contributions from unwanted redshift ranges.
- It reorganizes weak-lensing data so that only specific redshift ranges contribute to the signal.
- BNT aligns angular (ℓ) and physical (k) scales.
- This could help mitigate baryonic effects by optimally removing sensitivity to poorly modeled small scales and controlling scale leakage.
BNT maps
no BNT
BNT
How are statistics impacted?
Power Spectrum
l1-norm
* This could help mitigate baryonic effects by optimally removing sensitivity to poorly modeled small scales and controlling scale leakage?
How about contours?
| Statistic | FoM |
|---|---|
| PS | 8 |
| Peaks | 10 |
| l1-norm | 19 |
| Statistic | Factor |
|---|---|
| PS | 12.6 |
| Peaks | 22.9 |
| l1-norm | 15.3 |
Loss of power compared to contours without cuts
| Statistic | Factor |
|---|---|
| PS | 6.6 |
| Peaks | 6.3 |
| l1-norm | 7.0 |
Loss of power compared to contours with cuts
Conclusions
Part 2
- Unmodeled baryonic effects can lead to significant bias in cosmological parameters.
- To remove the bias from baryons, stringent scale cuts are required (ℓ~500 ).
- Even after scale cut at ℓ~500, HOS maintain significantly increased constraining power, compared to the PS.
- The BNT transform, while offering more physical scale cuts, leads to a significant inflation in contour size.
Part 1
- Mass mapping can have a significant impact on cosmological contours
- With a state-of-the-art mass-mapping method (MCALens) we managed to get 2.6x improvement in FoM over KS.