PageSourceSearch

https://wuzihao.site/research/

html wuzihao.site collected 2026-10-09 22:09:21 UTC 13,887 bytes, 346 lines download raw bytes

1<!doctype html>
2<!--
3  Minimal Mistakes Jekyll Theme 4.26.2 by Michael Rose
4  Copyright 2013-2024 Michael Rose - mademistakes.com | @mmistakes
5  Free for personal and commercial use under the MIT license
6  https://github.com/mmistakes/minimal-mistakes/blob/master/LICENSE
7-->
8
9<html lang="en-US" class="no-js">
10  <head>
11    <meta charset="utf-8">
12
13<!-- begin _includes/seo.html --><title>Research  |  Zihao Wu</title>
14<meta name="description" content="Zihao Wu is a PhD student in Astronomy at Harvard studying galaxies in the first 500 Myr with JWST.">
15
16
17  <meta name="author" content="Zihao Wu">
18  
19
20
21<meta property="og:type" content="website">
22<meta property="og:locale" content="en_US">
23<meta property="og:site_name" content="Zihao Wu">
24<meta property="og:title" content="Research">
25<meta property="og:url" content="https://zihaowu-astro.github.io/research/">
26
27
28  <meta property="og:description" content="Zihao Wu is a PhD student in Astronomy at Harvard studying galaxies in the first 500 Myr with JWST.">
29
30
31
32
33
34
35
36
37
38
39
40
41<link rel="canonical" href="https://zihaowu-astro.github.io/research/">
42
43
44
45
46
47
48
49
50
51
52
53
54<!-- end _includes/seo.html -->
55
56
57
58  <link href="/feed.xml" type="application/atom+xml" rel="alternate" title="Zihao Wu Feed">
59<meta name="viewport" content="width=device-width, initial-scale=1.0">
60
61<script>
62  document.documentElement.className = document.documentElement.className.replace(/\bno-js\b/g, '') + ' js ';
63  
64</script>
64
65
66<!-- For all browsers -->
67<link rel="stylesheet" href="/assets/css/main.css">
68<link rel="preload" href="https://cdn.jsdelivr.net/npm/@fortawesome/fontawesome-free@latest/css/all.min.css" as="style" onload="this.onload=null;this.rel='stylesheet'">
69<noscript><link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/@fortawesome/fontawesome-free@latest/css/all.min.css"></noscript>
70
71
72
73    <!-- start custom head snippets -->
74
75<!-- insert favicons. use https://realfavicongenerator.net/ -->
76
77<!-- end custom head snippets -->
78
79<!-- favicons -->
80
81
82  </head>
83
84  <body class="layout--splash wide-figures" dir="ltr">
85    <nav class="skip-links">
86  <ul>
87    <li><a href="#site-nav" class="screen-reader-shortcut">Skip to primary navigation</a></li>
88    <li><a href="#main" class="screen-reader-shortcut">Skip to content</a></li>
89    <li><a href="#footer" class="screen-reader-shortcut">Skip to footer</a></li>
90  </ul>
91</nav>
92
93    
94
95<div class="masthead">
96  <div class="masthead__inner-wrap">
97    <div class="masthead__menu">
98      <nav id="site-nav" class="greedy-nav">
99        
100        <a class="site-title" href="/">
101           
102          
103        </a>
104        <ul class="visible-links"><li class="masthead__menu-item">
105              <a
106                href="/"
107                
108                
109              >Home</a>
110            </li><li class="masthead__menu-item">
111              <a
112                href="/assets/cv.pdf"
113                
114                
115              >CV</a>
116            </li><li class="masthead__menu-item">
117              <a
118                href="/research/"
119                
120                
121              >Research</a>
122            </li><li class="masthead__menu-item">
123              <a
124                href="/code/"
125                
126                
127              >Code</a>
128            </li><li class="masthead__menu-item">
129              <a
130                href="/moments/"
131                
132                
133              >Photos</a>
134            </li></ul>
135        
136        <button class="greedy-nav__toggle hidden" type="button">
137          <span class="visually-hidden">Toggle menu</span>
138          <div class="navicon"></div>
139        </button>
140        <ul class="hidden-links hidden"></ul>
141      </nav>
142    </div>
143  </div>
144</div>
145
146
147    <div class="initial-content">
148      
149
150<div id="main" role="main">
151  <article class="splash" itemscope itemtype="https://schema.org/CreativeWork">
152    <meta itemprop="headline" content="Research">
153    
154    
155    
156
157    <section class="page__content" itemprop="text">
158      <h2 id="an-earliest-overdensity-structure-at-cosmic-dawn">An Earliest Overdensity Structure at Cosmic Dawn</h2>
159
160<div class="entry-row">
161  <div class="entry-text">
162
163    <p>Wu et al. (2026) <a href="https://iopscience.iop.org/article/10.3847/1538-4357/ae7723"><em>The Astrophysical Journal</em>, 1006, 18</a></p>
164
165    <p>We discover a candidate galaxy overdensity at <em>z ≈ 10.5</em>, only 500 Myr after the Big Bang: 18 galaxies with consistent photometric redshifts within 8 comoving Mpc, at four times the field density. The structure hosts 30% of the bright galaxies and 50% of the star formation at <em>10 &lt; z &lt; 12</em> in JADES GOODS-S, which suggests that early luminous galaxies assemble preferentially in overdense environments. Its abundance significantly exceeds the expectation from cosmological simulations such as TNG100, pointing to more efficient galaxy and structure formation than current models predict.</p>
166
167    <p>We have been approved for a JWST/NIRSpec program in Cycle 5 to confirm the structure spectroscopically and to probe its ionized bubble through Lyα emission, and for an ALMA program in Cycle 13 to constrain the halo mass and the gas properties of its members.</p>
168
169  </div>
170
171  <div class="entry-fig">
172
173    <figure class=""><img src="/assets/fig/overdensity.webp" alt="image" /><figcaption>
174      Galaxy candidates at 10 &lt; z_phot &lt; 12 in the JADES/NIRCam GOODS–S footprint, color-coded by photometric redshift. Large circles are bright sources (F356W &gt; 8 nJy); contours trace the density field from the mean to 4σ above it. The overdensity sits on the west side of the field, peaking above four times the mean. North is up, east is to the left.
175
176    </figcaption></figure>
177
178  </div>
179</div>
180
181<h2 id="extreme-deep-observation-of-a-galaxy-at-z14">
181Extreme Deep Observation of a Galaxy at z=14</h2>
182
183<div class="entry-row">
184  <div class="entry-text">
185
186    <p>Wu et al. (2025) <a href="https://iopscience.iop.org/article/10.3847/1538-4357/ae01a1"><em>The Astrophysical Journal</em>, 992, 212</a></p>
187
188    <p>We lead the analysis of ultra-deep JWST observations of JADES-GS-z14-1 (<em>z = 13.86</em>), one of the two most distant galaxies known and by far the fainter of them. We found that it falls in a small overlap of JADES MIRI pointings and so carries a fortuitous 70-hour exposure, the deepest MIRI observation of any high-redshift galaxy. With 56 new hours of NIRSpec/PRISM spectroscopy and 16-band NIRCam imaging, the dataset probes a sub-luminous galaxy in the early Universe at a depth normally reserved for the brightest ones.</p>
189
190    <p>Working at this signal-to-noise required new machinery: exposure-level model-fitting photometry for NIRCam and MIRI, and a covariance-aware treatment of the PRISM spectra. The galaxy shows only weak metal lines, unlike the luminous <em>z &gt; 10</em> population, and departs from the expected size–line strength relation. The most natural reading is low metallicity together with a high escape fraction — which would make faint galaxies like this one efficient contributors to reionization while still early in their chemical enrichment.</p>
191
192  </div>
193  <div class="entry-fig">
194
195    <figure class=""><img src="/assets/fig/mosaic.jpg" alt="image" /><figcaption>
196      NIRCam and MIRI/F770W imaging of JADES-GS-z14-1, unresolved in every band. The 70-hour F770W exposure gives a rest-frame optical flux density of 7.9 ± 2.6 nJy (29.2 mag).
197
198    </figcaption></figure>
199
200    <figure class=""><img src="/assets/fig/spectrum.webp" alt="spectrum" /><figcaption>
201      NIRSpec/PRISM spectrum of JADES-GS-z14-1. The 56-hour exposure shows no emission lines, but a tentative — and puzzling — absorption feature at 2µm.
202
203    </figcaption></figure>
204
205  </div>
206</div>
207
208<h2 id="detection-of-oxygen-emission-from-a-galaxy-at-z14">Detection of Oxygen Emission from a Galaxy at z=14</h2>
209
210<div class="entry-row">
211  <div class="entry-text">
212
213    <p>Helton et al. (2025) <a href="https://www.nature.com/articles/s41550-025-02503-z">Nature Astronomy, 1-12</a></p>
214
215    <p>The MIRI/F770W detection of JADES-GS-z14-0 (<em>z = 14</em>) is blended with a bright foreground galaxy, so the measurement that anchors its rest-frame optical emission is also the hardest one to make. We developed a forward-modeling approach that uses the high-resolution NIRCam imaging to model the source and the contaminant together, then performs exposure-level forced photometry with full propagation of uncertainties and covariances. This yields the F770W flux for the most luminous galaxy known at <em>z</em> &gt; 14.  This method can be applied to any MIRI photometry limited by blending.</p>
216
217  </div>
218  <div class="entry-fig">
219
220    <figure class=""><img src="/assets/fig/GS-z14-0.webp" alt="miri" /><figcaption>
221      MIRI/F770W imaging of JADES-GS-z14-0. The foreground galaxy 183349 severely contaminates the measurement.
222
223    </figcaption></figure>
224
225  </div>
226</div>
227
228<h2 id="subtraction-of-scattered-light-artifacts-in-jwst-imaging">Subtraction of Scattered-light Artifacts in JWST Imaging</h2>
229
230<div class="entry-row">
231  <div class="entry-text">
232
233    <p>Wu et al. (2026) <a href="https://iopscience.iop.org/article/10.3847/1538-3881/ae8254"><em>The Astronomical Journal</em>, 172, 105</a></p>
234
235    <p>Wisps are among the most severe scattered-light artifacts in JWST/NIRCam, contaminating nearly a quarter of the area in certain detectors, and their morphology changes from exposure to exposure. Therefore, subtraction with a single fixed template may leave noticeable residuals. We developed a data-driven method that models each wisp as a combination of components learned from the data by non-negative matrix factorization (NMF), which captures the morphological variation of wisps. This method is part of the JADES data reduction pipeline and is released as the <a href="/code/">NMFwisp</a> package.</p>
236
237  </div>
238  <div class="entry-fig">
239
240    <figure class="align-center"><img src="/assets/fig/wisp_model.gif" alt="wisp" /><figcaption>
241      Wisp removal in JWST/NIRCam: the default STScI pipeline correction compared with the NMF-based method.
242
243    </figcaption></figure>
244
245  </div>
246</div>
247
248<h2 id="probing-wandering-imbhs-with-quasar-microlensing">Probing Wandering IMBHs with Quasar Microlensing</h2>
249
250<div class="entry-row">
251  <div class="entry-text">
252
253    <p>Wu &amp; Ho (2025) <a href="https://iopscience.iop.org/article/10.3847/1538-4357/adcec3"><em>The Astrophysical Journal 985, 2</em></a></p>
254
255    <p>Wandering intermediate-mass black holes are essentially invisible on their own, which leaves their population almost unconstrained. We show that microlensing of background quasars offers a way to find them, and that the signal is within reach of the 10-year LSST survey, especially for IMBHs hosting compact star cluster
255s. The lensing magnifies only the emission from scales smaller than the broad-line region, so a lensed quasar should present as a high-Eddington, weak-line source; cosmological peculiar motions then drive slow variability whose amplitude is the same in the ultraviolet, optical, and X-ray. IMBHs embedded in companion clusters produce complex caustic networks and correspondingly distinctive light curves. Each of these signatures is testable against the survey data.</p>
256
257  </div>
258  <div class="entry-fig">
259
260    <figure class=""><img src="/assets/fig/IMBH.webp" alt="imbh" /><figcaption>
261      Caustic network of a 1000 M⊙ black hole at <em>z</em> = 1, embedded in a stellar cluster of 30 stars pc⁻².
262
263    </figcaption></figure>
264
265  </div>
266</div>
267
268<h2 id="origin-of-agn-radio-jets-in-disk-galaxies">Origin of AGN Radio Jets in Disk Galaxies</h2>
269
270<div class="entry-row">
271  <div class="entry-text">
272
273    <p>Wu et al. (2022) <a href="https://iopscience.iop.org/article/10.3847/1538-4357/ac9cd5#back-to-top-target"><em>The Astrophysical Journal 941, 95</em></a></p>
274
275    <p>Large-scale radio jets are found almost exclusively in giant elliptical galaxies, and it has long been unclear whether morphology itself is what matters. We studied the rare disk galaxies that do host double radio lobes, drawn from the Gems of the Galaxy Zoos project, and found that the apparent preference for ellipticals is better explained as a mass bias: these radio-loud disks carry stellar masses an order of magnitude above typical disks, comparable to giant ellipticals. That two rare properties — extreme stellar mass and a powerful jet — coincide so consistently suggests jet formation is set by mass, and more fundamentally by black hole mass, rather than by the morphology of the host.</p>
276
277  </div>
278  <div class="entry-fig">
279
280    <figure class=""><img src="/assets/fig/radio.webp" alt="radio" /><figcaption>
281      HST image of a radio galaxy with clear disk morphology; white contours show VLA synchrotron emission from the jets.
282
283    </figcaption></figure>
284
285  </div>
286</div>
287
288    </section>
289  </article>
290</div>
291
292      
293    </div>
294
295    
296
297    <div id="footer" class="page__footer">
298      <footer>
299        <!-- start custom footer snippets -->
300
301<!-- end custom footer snippets -->
302        <!-- <div class="page__footer-follow">
303  <ul class="social-icons">
304    
305
306    
307
308    
309      <li><a href="/feed.xml"><i class="fas fa-fw fa-rss-square" aria-hidden="true"></i> Feed</a></li>
310    
311  </ul>
312</div> -->
313
314<div class="page__footer-copyright">&copy; 2026 <a href="https://zihaowu-astro.github.io">Zihao Wu</a>. Powered by <a href="https://jekyllrb.com" rel="nofollow">Jekyll</a> &amp; <a href="https://mademistakes.com/work/jekyll-themes/minimal-mistakes/" rel="nofollow">Minimal Mistakes</a>.</div>
315
316      </footer>
317    </div>
318
319    
320  
320<script src="/assets/js/main.min.js"></script>
320
321
322
323
324
325
326
327
328  <!-- start custom analytics snippet -->
329
330<script
331  data-goatcounter="https://zihaowu.goatcounter.com/count"
332  async
333  src="https://gc.zgo.at/count.js"></script>
333
334
335
336<!-- end custom analytics snippet -->
337
338
339
340
341
342
343
344
345  </body>
346</html>

Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.