Differences

This shows you the differences between two versions of the page.

--- courses:ast403:redshift-surveys [2026/04/06 09:16] – shuvo
+++ courses:ast403:redshift-surveys [2026/04/06 22:27] (current) – [DESI BAO Survey] shuvo
@@ Line 12: / Line 12: @@
 The instrument possessed a 2-degree field of view on the sky (roughly four times the diameter of the full moon) and utilized a robotic arm to position 400 optical fibers onto the focal plane. Each fiber was aligned with a pre-selected target galaxy. This allowed astronomers to capture the spectra—and thus the redshifts—of 400 galaxies simultaneously in a single observation.
+See how it works: [[https://www.youtube.com/watch?v=qOtkI4HYzTk]]
 **Survey Scope and Geometry:**\\
@@ Line 30: / Line 32: @@
 //1.  Mapping the Cosmic Web:// The 2dF slice maps provided stunning visual confirmation of the "cosmic web." Galaxies were not distributed uniformly, nor were they clumped randomly; they formed a vast network of dense nodes (superclusters), long connecting filaments, and vast, empty voids. \\
 //2.  Matter Density:// By analyzing the way galaxies clustered and incorporating redshift-space distortions (the apparent squashing of galaxy clusters due to their peculiar velocities), the 2dF team precisely constrained the total matter density parameter of the universe, $\Omega_m$, finding it to be roughly 30% of the critical density.\\
-**3.  Upper Limit on Neutrino Mass:** The survey placed some of the first stringent cosmological limits on the total mass of neutrino species, as massive neutrinos would "free-stream" in the early universe and wash out the formation of structure on small scales.
+//3.  Upper Limit on Neutrino Mass:// The survey placed some of the first stringent cosmological limits on the total mass of neutrino species, as massive neutrinos would "free-stream" in the early universe and wash out the formation of structure on small scales.
+[{{ :courses:ast403:2df_slice_blue_big.gif?600 | Fig 1: 2dF redshift map.}}]
 **2dFGRS and Baryon Acoustic Oscillations:**\\
@@ Line 38: / Line 43: @@
 This detection proved that the sound waves generated in the primordial plasma (seen in the Cosmic Microwave Background at $z \approx 1100$) had successfully translated into the physical clustering of galaxies billions of years later at $z \approx 0.11$.
+[{{ :courses:ast403:2df_power.jpg?600 | Fig 2: Matter power spectrum.}}]
+===== SDSS BAO Survey =====
+While the 2dFGRS provided one of the first co-detections of the Baryon Acoustic Oscillation (BAO) signal, it was the Sloan Digital Sky Survey (SDSS) that transformed BAO from a fascinating theoretical prediction into the premier tool for precision cosmology. Operating out of the Apache Point Observatory in New Mexico, SDSS is a multi-generational project that has systematically mapped the universe over the last two decades.
+For graduate students studying large-scale structure, understanding the progression of SDSS is crucial, as its distinct phases introduced new observational techniques and targeted different cosmic epochs.
+About SDSS: https://www.youtube.com/watch?v=UD6cOMpJlZU
+**The First Detection: SDSS-I and II (2000–2008):**
+The initial phases of SDSS utilized a 2.5-meter telescope equipped with a 640-fiber spectrograph. In 2005, alongside the 2dFGRS announcement, the SDSS team (led by Daniel Eisenstein) published a definitive detection of the BAO peak in the configuration space two-point correlation function, $\xi(r)$.
+To achieve this, SDSS targeted a specific class of objects called Luminous Red Galaxies (LRGs). LRGs are massive, old, elliptical galaxies. They are highly luminous, meaning they can be seen at great distances. More importantly, they reside in the most massive dark matter halos, meaning they are highly "biased" tracers of the underlying matter distribution ($b > 2$). This amplifies the clustering signal, making the BAO peak easier to detect against the statistical noise.
+The 2005 detection localized the BAO peak at a comoving separation of approximately $100\ h^{-1}$ Mpc (equivalent to $r_s \approx 147$ Mpc) using a sample of about 46,000 LRGs out to $z \approx 0.47$.
+[{{ :courses:ast403:sdss_bao.webp?600 | Fig 3: SDSS BAO feature.}}]
+**The Baryon Oscillation Spectroscopic Survey (SDSS-III, 2009–2014):**
+Recognizing the power of BAO as a dark energy probe, SDSS-III launched BOSS, a survey explicitly designed to measure the BAO scale to an unprecedented 1% precision. The spectrograph was upgraded to 1,000 fibers to drastically increase the survey speed.
+BOSS tackled the BAO measurement on two distinct frontiers:
+//1.  The Low-Redshift Universe ($z < 0.7$):// BOSS measured spectra for 1.5 million LRGs, vastly increasing the volume mapped by SDSS-I/II and providing incredibly tight constraints on the angular diameter distance $D_A(z)$ and the Hubble parameter $H(z)$.
+**2.  The High-Redshift Universe ($z > 2.1$):** At redshifts beyond $z \sim 1$, galaxies become too faint to observe efficiently in massive numbers. BOSS pioneered a revolutionary technique: using the Lyman-$\alpha$ Forest of distant quasars.
+**The Lyman-$\alpha$ Forest Technique:**\\
+Instead of using galaxies as discrete point masses, BOSS observed distant, highly luminous quasars. As the light from a quasar travels toward Earth, it passes through intervening clouds of neutral hydrogen gas. Each cloud absorbs a specific fraction of the quasar's light at the Lyman-$\alpha$ transition wavelength ($121.6$ nm in the rest frame of the cloud). Because the clouds are at different redshifts, this absorption creates a "forest" of lines in the quasar's spectrum.
+The transmitted flux fraction, $F$, is related to the optical depth, $\tau$, of the hydrogen gas:
+$$F = \exp(-\tau)$$
+By measuring the correlation function of the flux transmission $\xi_F(r)$ between thousands of different quasar sightlines, BOSS successfully detected the BAO scale in the continuous intergalactic medium at $z \approx 2.5$, an epoch before dark energy dominated the universe.
+**eBOSS: Extended BOSS (SDSS-IV, 2014–2020)**\\
+While BOSS mapped the low-redshift (galaxy) and high-redshift (Lyman-$\alpha$) universe, it left a "redshift desert" in the middle ($0.6 < z < 2.2$). This epoch is critical, as it is exactly when the universe transitioned from matter-dominated deceleration to dark-energy-dominated acceleration.
+eBOSS was designed to fill this gap using new tracers of the matter density field:\\
+* **Emission Line Galaxies (ELGs):** Young, star-forming galaxies that exhibit strong emission lines (like [O II]). These are easier to detect at intermediate redshifts than the passive LRGs.\\
+* **Quasars as Point Sources:** In addition to using quasars as backlights for the Lyman-$\alpha$ forest, eBOSS used the quasars themselves as discrete tracers of the matter distribution to map the $0.8 < z < 2.2$ regime.\\
+[{{ :courses:ast403:sdss_ba01.png?600 | Fig 4: SDSS BOSS BAO detection.}}]
+**BAO Reconstruction:**\\
+A vital mathematical and computational advancement popularized during the SDSS BOSS era is BAO Reconstruction.
+Over billions of years, non-linear gravitational collapse and the peculiar velocities of galaxies (bulk flows) smear out the primordial BAO peak in the correlation function, degrading the precision of the standard ruler.
+Reconstruction attempts to reverse this cosmic evolution. By taking the observed smoothed density field $\delta(\mathbf{x})$, physicists use the Zel'dovich approximation to estimate the displacement field $\boldsymbol{\Psi}$ that moved galaxies from their initial Lagrangian positions $\mathbf{q}$ to their observed Eulerian positions $\mathbf{x}$:
+$$\mathbf{x} = \mathbf{q} + \boldsymbol{\Psi}(\mathbf{q})$$
+By moving the observed galaxies backward along these estimated displacement vectors, the non-linear smearing is partially undone. This process drastically sharpens the BAO peak in $\xi(r)$, recovering lost information and significantly improving the statistical constraints on cosmological parameters.
+===== DESI BAO Survey =====
+As we move to the present day in our lecture, we enter the era of "Stage-IV" dark energy experiments. The current flagship ground-based BAO experiment is the Dark Energy Spectroscopic Instrument (DESI), which represents a massive leap in survey speed and mapping volume over its predecessors like SDSS and BOSS.
+Operating from the Mayall 4-meter telescope at Kitt Peak National Observatory in Arizona, DESI began its main survey in 2021. Its primary objective is to measure the expansion history of the universe with exquisite precision to constrain the nature of Dark Energy.
+**The Instrument: A Marvel of Engineering:**\\
+The leap in efficiency achieved by DESI comes from its focal plane. Unlike older surveys that required astronomers to manually plug optical fibers into custom-drilled metal plates for each observation, DESI features a focal plane packed with 5,000 robotic fiber positioners.
+These automated micromotors can reconfigure the arrangement of the 5,000 fibers in a matter of minutes, aligning them with new target galaxies to an accuracy of a few microns. The light from these fibers is fed into ten spectrographs, each split into three spectral bands (blue, red, and near-infrared). This allows DESI to capture the spectra of 5,000 targets simultaneously, enabling it to map tens of millions of galaxies in just five years.
+DESI focal plane: https://www.youtube.com/watch?v=g1LVMox0KNc&t=50s
+**The Tracers: A Multi-Target Strategy:**\\
+To map the universe continuously from the local volume out to the era of matter domination, DESI utilizes a multi-target strategy, targeting different classes of objects at different redshifts:\\
+//1.  Bright Galaxy Sample (BGS):// Probes the low-redshift universe ($z < 0.4$).\\
+//2.  Luminous Red Galaxies (LRGs):// Extends from $0.4 < z < 0.8$.\\
+//3.  Emission Line Galaxies (ELGs)// Star-forming galaxies mapped from $0.8 < z < 1.6$, representing the bulk of DESI's target catalog.\\
+//4.  Quasars (QSOs):// Used both as discrete density tracers ($1.6 < z < 2.1$) and as backlights for the **Lyman-$\alpha$ forest** ($z > 2.1$).
+**DESI Year 1 (Y1) Results and Cosmological Constraints:**
+In April 2024, the DESI collaboration released its Year 1 (Y1) cosmology results, analyzing a map comprising over 6 million extragalactic objects—already the largest and most precise 3D map of the universe ever constructed.
+The DESI Y1 BAO measurements successfully isolated the comoving sound horizon scale across seven distinct redshift bins ranging from $z = 0.1$ to $z = 4.2$.
+On its own, the DESI BAO data proved highly consistent with a spatially flat $\Lambda$CDM universe, yielding a tightly constrained matter density parameter:
+$$\Omega_m = 0.295 \pm 0.015$$
+When combined with Cosmic Microwave Background (CMB) data from the Planck satellite, the derived Hubble constant is measured at:
+$$H_0 = 67.97 \pm 0.38 \text{ km s}^{-1} \text{ Mpc}^{-1}$$
+**The Evolving Equation of State: A Hint of New Physics?**\\
+The most highly discussed outcome of the DESI Y1 data release pertained to the equation of state of Dark Energy, denoted as $w$. In the standard $\Lambda$CDM model, dark energy is a cosmological constant ($\Lambda$), meaning its energy density does not change as the universe expands. For a cosmological constant, the equation of state is exactly:
+$$w = \frac{p}{\rho c^2} = -1$$
+However, cosmologists test alternative models where dark energy is dynamic (e.g., quintessence), meaning its density evolves over time. To test this, researchers often employ the **Chevallier-Polarski-Linder (CPL) parameterization**, which models the equation of state as a function of the scale factor $a$ (where $a = 1/(1+z)$):
+$$w(a) = w_0 + w_a (1 - a)$$
+Here:\\
+* $w_0$ is the current value of the dark energy equation of state at $z = 0$.\\
+* $w_a$ dictates how quickly $w$ evolves with time.\\
+* (Note: $\Lambda$CDM strictly requires $w_0 = -1$ and $w_a = 0$).\\
+When the DESI collaboration combined their Y1 BAO measurements with CMB data and Type Ia Supernova data, the statistical fit preferred a parameter space where **$w_0 > -1$ and $w_a < 0$**.
+Depending on the specific supernova dataset used (e.g., Pantheon+, Union3, or DES-SN5YR), the combined data deviated from the standard $\Lambda$CDM prediction by roughly **$2.5\sigma$ to $3.9\sigma$**.
+While this does not meet the $5\sigma$ threshold required to officially claim a "discovery" of new physics, it provided the cosmological community with a tantalizing hint that dark energy might be weakening over cosmic time. Future data releases from DESI (spanning Years 3 and 5) will determine whether this deviation is a statistical fluctuation, an unrecognized systematic error, or the first definitive crack in the $\Lambda$CDM model.
+[{{ :courses:ast403:desi_bao1.jpg?600 | FFig 5: DESI BAO detection.}}]