hep-th0607120/mt.tex
1: \documentclass[12pt]{article}
2: \usepackage{cite,graphicx,amsmath,amssymb,bbm}
3: \usepackage{epsf}
4: 
5: \newcommand{\be}{\begin{equation}}  
6: \newcommand{\ee}{\end{equation}}  
7: \newcommand{\bea}{\begin{eqnarray}}  
8: \newcommand{\eea}{\end{eqnarray}}  
9: \newcommand{\UV}{{\rm UV}}
10: \newcommand{\IR}{{\rm IR}}
11: \renewcommand{\Re}{\operatorname{Re}}
12: \renewcommand{\Im}{\operatorname{Im}}
13: \newcommand{\ol}[1]{\overline{#1}}
14: 
15: \addtolength\topmargin{-50pt}
16: \addtolength\textheight{105pt}
17: \addtolength\textwidth{60pt}
18: \addtolength\oddsidemargin{-38pt}
19: \setlength{\parindent}{20pt}
20: \setlength{\parskip}{6pt}
21: \frenchspacing
22: \sloppy
23: 
24: \begin{document}
25: 
26: \thispagestyle{empty}
27: \vspace*{.2cm}
28: \noindent
29: HD-THEP-06-12 \hfill 18 July 2006
30: \\
31: \noindent
32: OUTP-DR-06 01P
33: 
34: \vspace*{1.0cm}
35: 
36: \begin{center}
37: {\Large\bf The Ubiquitous Throat}
38: \\[1.5cm]
39: {\large A.~Hebecker$\,^a$, and J.~March-Russell$\,^b$}\\[.5cm]
40: {\it ${}^a$ Institut f\"ur Theoretische Physik, Universit\"at Heidelberg,
41: Philosophenweg 16 und 19\\ D-69120 Heidelberg, Germany}
42: \\[.3cm]
43: {\it ${}^b$ Rudolf Peierls Centre for Theoretical Physics, University 
44: of Oxford, 1 Keble Road\\ Oxford OX1 3NP, UK}
45: \\[.4cm]
46: {\small\tt (a.hebecker@thphys.uni-heidelberg.de} {\small and}
47: {\small\tt j.march-russell1@physics.ox.ac.uk\,)}
48: \\[1.0cm]
49: 
50: {\bf Abstract}
51: \end{center} 
52: We attempt to quantify the widely-held belief that large hierarchies 
53: induced by strongly-warped geometries are common in the string theory
54: landscape. To this end, we focus on the arguably best-understood subset 
55: of vacua -- type IIB Calabi-Yau orientifolds with non-perturbative K\"ahler 
56: stabilization and a SUSY-breaking uplift (the KKLT setup). Within this 
57: framework, vacua with a realistically small cosmological constant are 
58: expected to come from Calabi-Yaus with a large number of 3-cycles. For 
59: appropriate choices of flux numbers, many of these 3-cycles can, in general, 
60: shrink to produce near-conifold geometries. Thus, a simple statistical 
61: analysis in the spirit of Denef and Douglas allows us to estimate the 
62: expected number and length of Klebanov-Strassler throats in the given 
63: set of vacua. We find that throats capable of explaining the electroweak 
64: hierarchy are expected to be present in a large fraction of the landscape 
65: vacua while shorter throats are essentially unavoidable in a statistical 
66: sense.
67: 
68: 
69: 
70: 
71: 
72: \newpage
73: \section{Introduction}
74: Since the seminal work of Giddings, Kachru and Polchinski~\cite{
75: Giddings:2001yu} following on from the foundational papers 
76: of Refs.\cite{earlywarp} and \cite{Verlinde:1999fy},
77: it has become common knowledge that strongly warped 
78: regions or throats are a natural feature of type IIB flux compactifications
79: (see~\cite{Grana:2005jc} for a recent review). Moreover, thanks to the KKLT 
80: construction~\cite{Kachru:2003aw}, the very same class of models has become 
81: the nucleus of the large and growing collection of metastable de-Sitter 
82: vacua of string theory (known with a varying degree of rigour) which are 
83: generally referred to as the `string theory landscape' \cite{landscape}. Following the 
84: line of thought developed by Douglas and collaborators~\cite{Douglas:2003um,
85: Denef:2004ze,Denef:2004cf}, it is then natural to attempt to link the 
86: presence of throats quantitatively to the assumption that we live in one of 
87: the numerous type IIB orientifold models with 3-form flux. It is the aim of 
88: the present paper to understand to which extent throat and multi-throat 
89: geometries can be considered a prediction of the type IIB landscape proposal. 
90: 
91: To be specific, we will focus on the oldest and arguably simplest 
92: situation~\cite{Kachru:2003aw} in which, given a model where all 
93: complex structure moduli are stabilized by 3-form flux, the single K\"ahler 
94: modulus is stabilized non-perturbatively by gaugino condensation or D3-brane 
95: instantons. We have every reason to expect that our conclusions, which will 
96: mainly be related to the distribution of 3-form flux quanta on the various 
97: 3-cycles, remain valid if K\"ahler moduli are stabilized by the interplay of 
98: perturbative and nonperturbative physics~\cite{Balasubramanian:2004uy, 
99: Balasubramanian:2005zx} or even in an entirely perturbative 
100: fashion~\cite{Saltman:2004sn,vonGersdorff:2005bf}. Similarly, we do not 
101: expect our conclusions to be affected by the modifications and extensions 
102: of the stabilization mechanism required in situations with more than one 
103: K\"ahler modulus (see e.g.~\cite{Balasubramanian:2005zx}). 
104: 
105: Given that all geometric moduli are stabilized in a supersymmetric AdS 
106: vacuum as described above, we assume, following KKLT~\cite{Kachru:2003aw}, 
107: that a small supersymmetry breaking effect, such as the presence of 
108: anti-D3-branes in one of the warped regions, uplifts this vacuum to a 
109: de-Sitter vacuum with realistic cosmological constant. We choose to focus on 
110: this (by now classic) scenario since the metastability of such uplifted vacua 
111: is essentially guaranteed in the limit of a parametrically small AdS 
112: cosmological constant before the uplift. We will comment on this in more 
113: detail below. However, we emphasize again that our decision to be so 
114: restrictive in our choice of models is motivated solely be the desire to 
115: keep the non-essential parts of our analysis short and simple. We expect 
116: that the distribution of throats emerging from our analysis will be similar 
117: in a much wider class of flux vacua. 
118: 
119: Given the above considerations, we focus on the distribution of throats in 
120: type IIB Calabi-Yau orientifolds with a large number of 3-cycles and under
121: the restriction that the total flux superpotential $W_0$ at the SUSY minimum 
122: is parametrically small. It is natural to expect that, as a result of the 
123: random choice of a large number of independent flux quanta for the various 
124: 3-cycles, some of these 3-cycles will automatically carry only a small
125: number of flux. If this occurs for a 3-cycle that can shrink to produce 
126: a conifold singularity~\cite{Candelas:1989js} (which may be the generic 
127: situation) and if the flux carried by the dual cycle is not small, a 
128: Klebanov-Strassler throat~\cite{Klebanov:1998hh,Klebanov:2000hb} with 
129: an exponentially large hierarchy develops. This is the naive expectation and 
130: at the same time our main result: The more detailed analysis described 
131: in bulk of the paper confirms that one has to expect large hierarchies of 
132: scales~\cite{Randall:1999ee,Verlinde:1999fy,Goldberger:1999uk} and multiple 
133: throats~\cite{smallnumbers,Barnaby:2004gg,Cascales:2003wn} in generic 
134: orientifold models of the landscape. 
135: 
136: At a more technical level, we will replace the above heuristic argument 
137: about `accidentally' small 3-cycles by the quantitatively well-known fact
138: that vacua accumulate near conifold points~\cite{Denef:2004ze,
139: Giryavets:2004zr,Conlon:2004ds,Eguchi:2005eh}. If many such conifold points 
140: are present in the moduli space of a given Calabi-Yau, the probability of 
141: being far away from any of them becomes extremely small. In this sense, the 
142: presence of throats becomes a prediction of the given branch of the string 
143: theory landscape.
144: 
145: 
146: 
147: 
148: 
149: \section{The relevant set of vacua}
150: Following~\cite{Denef:2004ze}, we consider the orientifold limit of an 
151: F-theory compactification based on a four-fold with Euler number $\chi_4$. 
152: The 3-form flux on this orientifold can be quantified by a flux vector
153: $N\in {\mathbb Z}^{2K}$. Its dimension is given by $2K=4(h^{2,1}_-+1)$, where 
154: $h^{2,1}_-$ is the number of complex structure moduli.\footnote{
155: The 
156: index `$-$' is used since, as an alternative to the F theory construction, 
157: one may think of orientifolding a smooth Calabi-Yau 3-fold to obtain a
158: given model. The relevant cycles are those which are odd under the 
159: orientifold projection. Note that our $K$ is $K/2$ in the notation 
160: of~\cite{Denef:2004ze}
161: }
162: Allowing for a contribution $N_{D3}$ to the total D3-brane charge from 
163: freely moving D3 branes $(N_{D3}>0)$ or anti-D3-branes $(N_{D3}=-N_{\ol{D3}} 
164: <0)$, the tadpole cancellation condition reads
165: \be
166: \frac{\chi_4}{24}=\frac{1}{2}N^T\Sigma N+N_{D3}\,,
167: \qquad\mbox{where}\qquad
168: \Sigma\equiv\left(\begin{array}{cc}0&\mathbbm{1}\\ \mathbbm{1}&
169: 0\end{array}\right)\,.
170: \ee
171: Assuming that the orientifold planes of the model preserve the same 
172: supersymmetry as D3 (rather than anti-D3) branes and focusing on 
173: supersymmetric vacua,\footnote{
174: To 
175: be more precise, these are SUSY-breaking no scale vacua which turn into 
176: supersymmetric AdS vacua once non-perturbative K\"ahler stabilization is 
177: taken into account.
178: }
179: one requires $N_{D3}>0$. The flux vector is then subject to the constraint 
180: \be 
181: \frac{1}{2}N^T\Sigma N\equiv L\le L_* \equiv \frac{\chi_4}{24}\,.
182: \ee
183: The number of SUSY vacua available in this situation was estimated in~\cite{ 
184: Denef:2004ze} to be 
185: \be
186: {\cal N}_{susy}(L\le L_*) \sim \frac{L_*^K}{K!}\,.\label{nsusy}
187: \ee
188: 
189: We are, however, interested specifically in realistic vacua (i.e. vacua with 
190: small positive cosmological constant) originating from non-perturbative 
191: K\"ahler stabilization combined with an anti-D3-brane uplift based on a 
192: small positive $N_{\ol{D3}}$. To ensure stability, no freely moving D3 branes 
193: should be present in this construction. Furthermore, to guarantee perturbative 
194: control and a sufficiently long lifetime of the metastable anti-D3-brane 
195: configuration at the bottom of the throat, we require $N_{\ol{D3}}\leq 
196: N_{\ol{D3}\,,\,max}$. An upper bound on $N_{\ol{D3}\,,\,max}$ is provided by 
197: the classically-allowed decay process studied in~\cite{Kachru:2002gs} which 
198: limits the range of metastability to $N_{\ol{D3}}< 0.08 M$ where $M$ is the 
199: RR-flux quantum. As in our counting of flux vacua we scan over flux quanta up 
200: to $L_*$ we take the parametric dependence $N_{\ol{D3}\,,\,max}\ll L_*$. 
201: 
202: The number of such `uplifted' vacua can be estimated by an appropriate 
203: modification of Eq.~(\ref{nsusy}):
204: \be 
205: {\cal N}_{uplift}={\cal N}_{susy}(L_*<L\le L_*+N_{\ol{D3}\,,\,max}) 
206: \sim N_{\ol{D3}\,,\,max}\,\frac{L_*^{K-1}}{(K\!-\!1)!}\
207: \sim \frac{L_*^K}{K!}\,,
208: \ee
209: where we have Taylor expanded in $N_{\ol{D3}\,,\,max}$ and dropped irrelevant 
210: non-exponential factors in the last expression to simplify the final formula. 
211: Thus, ${\cal N}_{uplift}$ has the same parametric behaviour as ${\cal 
212: N}_{susy}$. Clearly, the cosmological constants of these uplifted vacua can 
213: have both signs and vary widely in their value. The source for this 
214: variation is the flux superpotential 
215: \be
216: W=\int G_3\wedge\Omega\,,
217: \ee
218: which provides a negative contribution $\sim |W_0|^2$ in each vacuum, to be 
219: (under- or over-) compensated by the uplift $\sim N_{\ol{D3}}$. Given that 
220: both Re$\,W_0$ and Im$\,W_0$ depend linearly on the flux vector, one expects 
221: a uniform distribution of vacua in the central region of the complex $W_0$ 
222: plane. This, in turn, implies a uniform distribution of $|W_0|^2$ on the 
223: positive real axis. If we ignore any moderate volume suppression and 
224: non-exponential factors depending on $L_*$ and $K$, the maximal size of 
225: $W_0$ is string scale (i.e. ${\cal O}(1)$ in our units). Thus, the 
226: probability that the negative contribution $\sim |W_0|^2$ compensates a 
227: fixed positive $V_{uplift}$ with enough precision to come close to the 
228: observed cosmological constant $\Lambda$ is approximately equal to $\epsilon 
229: \sim\Lambda 
230: \sim 10^{-120}$. ($V_{uplift}$ should be small enough to allow perturbative 
231: control but large enough to avoid any peculiarity that the $W_0$ 
232: distribution might have very close to the origin.) We conclude that one 
233: needs geometries with ${\cal N}_{uplift} \sim 10^{120}$ to have ${\cal O}(1)$ 
234: probability for a (cosmologically) realistic vacuum to exist. 
235: 
236: We are interested in an estimate for the lowest $K$ that is consistent within 
237: the present framework. Thus, we choose $L_*=\chi/24$ as large as possible 
238: (within the presently known set of Calabi-Yau 4-folds) and estimate $K$ on 
239: the basis of 
240: \be
241: \frac{L_*^K}{K!}\sim \frac{1}{\epsilon}\,.\label{keq}
242: \ee
243: Ignoring non-exponential factors in Stirling's formula and assuming 
244: that $\log(eL_*/K)\simeq {\cal O}(1)$, one finds $K_0\sim \log(1/\epsilon)$.
245: A better estimate of $K$ follows from replacing $K!$ with $(K_0/e)^K$ 
246: on the lhs of Eq.~(\ref{keq}) leading to
247: \be
248: K\sim \frac{\log(1/\epsilon)}{\log[eL_*/\log(1/\epsilon)]}\,.
249: \ee
250: For $L_*\sim 10^4$ (see, e.g.,~\cite{Klemm:1996ts}), we find $K\sim 60$,
251: corresponding to $h^{2,1}_-\sim 30$. 
252: 
253: It is important to keep in mind that this is just a lower bound and 
254: that, most probably, the number of cycles of a `typical' flux 
255: compactification with realistic cosmological constant is significantly 
256: larger. For example, one might say that with typical Calabi-Yaus having 
257: $h^{2,1}\sim 100...200$ (see e.g.~\cite{Candelas:1994hw,Avram:1996pj, 
258: Denef:2004dm})\footnote{
259: Extreme 
260: cases of $h^{2,1}\sim 500$ are known, see, eg, 
261: {\tt http://hep.itp.tuwien.ac.at/\~{}kreuzer/CY/}
262: }  
263: we can very naively expect that $h^{2,1}_{-}\sim 50...100$ is typical.
264: Moreover, the scaling of the number of vacua (as implied by Eq.~(\ref{keq}) 
265: for $K<L_*$) suggests that CY's with the largest
266: possible value of $K$ are exponentially preferred in that they allow a far 
267: greater number of flux vacua. For example, ${\cal N}_{susy}(K=200)/{\cal 
268: N}_{susy}(K=60) \sim 10^{270}$ for fixed $L_*\sim 10^4$.
269:  
270: Given our ignorance of the model describing our vacuum as well as of the 
271: mechanism choosing physical compactification manifolds, we keep 
272: $h^{2,1}_{-}$ and $\chi_4$ (or equivalently $K$ and $L_*$) as unknown 
273: parameters with the order of magnitude given above.
274: 
275: The complex structure moduli spaces of such complicated orientifold models 
276: have not been analyzed in detail. It is clear that they will contain various
277: regions where certain 3-cycles blow up or shrink to zero size. We will
278: henceforth ignore the former `large complex structure' regions although
279: they might, in fact, be interesting and important to study. Instead, in 
280: this paper we focus on the singularities arising when one or more of the 
281: 3-cycles shrink. We want to argue that, in many cases, these singularities 
282: are `nodes' or `ordinary double points', which are particularly common 
283: singularities of complex varieties. Nodal 3-folds arise naturally in 
284: algebraic topology, one of the prominent examples being the various 
285: singular limits of the quintic hypersurface in 4d complex projective space. 
286: In this specific case, it is known that the `generic' singular space has a
287: single node~\cite{Lefschetz}. From the perspective of the Calabi-Yau 3-fold 
288: defined in this way, such a point corresponds to a conifold 
289: singularity~\cite{Candelas:1989js} (which develops as one of the 3-cycles 
290: shrinks). Furthermore, a large set of smooth Calabi-Yaus is linked by 
291: conifold transitions into a `Web' (including, in particular, the 
292: quintic)~\cite{Candelas:1989ug}. In each case, the singular intermediate 
293: situation is approached from one side of the transition by the shrinking of 
294: a number of 3-cycles with $S^3$-topology (the conifold limit mentioned above)
295: \cite{Curio:2000sc}.  From this we conclude that the conifold limit is a common
296: (possibly the generic) way in which a 3-cycle of a Calabi-Yau shrinks. More
297: specifically, we assume in the following that an ${\cal O}(1)$ fraction of the possible 
298: limits of shrinking 3-cycles of the models under consideration correspond to 
299: conifold points. It is an interesting question (which goes beyond the scope 
300: of this work) to understand for how many of the known Calabi-Yau orientifolds 
301: this assumption holds. 
302: 
303: The distribution of flux vacua in the vicinity of such conifold 
304: points has been analyzed at least for certain simple examples. It has been 
305: found that vacua accumulate near these points. This can be understood 
306: intuitively by recalling that the distance $|z|$ from a conifold point is 
307: given by~\cite{Giddings:2001yu}
308: \be 
309: z\sim \exp(-2\pi P/g_sM)\,.
310: \ee
311: Here $z$ is the complex structure modulus corresponding to the shrinking 
312: 3-cycle while $M$ and $P$ are the numbers of flux quanta on the conifold 
313: cycle and on its dual. It is then clear that a smooth distribution of flux 
314: quanta can lead to a strong enhancement of the number of vacua with 
315: exponentially small $z$. 
316: 
317: More specifically, it was shown in~\cite{Denef:2004ze} that, in a given 
318: model with one conifold point at $z=0$ and a fixed tadpole constraint 
319: $L_*$, the fraction of vacua with conifold cycle smaller than $|z|$ decays 
320: as 
321: \be 
322: {\cal N}(z)\sim \frac{1}{\log(1/|z|)}\label{enh}
323: \ee
324: for $z\to 0$. Clearly, this implies an enhancement of the number of vacua 
325: very close to the conifold point relative to naive expectations that one 
326: might have on the basis of the canonical measure on ${\mathbb C}$. 
327: 
328: 
329: 
330: 
331: 
332: \section{Stability issues}
333: Before turning to our main interest, the distribution of throats, we would 
334: like to address the stability of the above set of vacua after uplifting. 
335: 
336: Naively, one might expect the following situation: We focus on the complex 
337: structure moduli $z_i$ and the dilaton modulus $\tau$. A generic flux 
338: induced modulus mass is ${\cal O}$(1) in string units (if we ignore any 
339: volume suppression $\sim{\cal O}(\mbox{few})$). Making the vacuum value 
340: $W_0$ of the superpotential $W$ parametrically small by tuning fluxes, we 
341: obtain a parametrically small cosmological constant $\Lambda_{\rm AdS}\sim 
342: -|W_0|^2$. Consider now the scalar potential
343: \be
344: V=e^K(K^{a\ol{b}}D_aW\ol{D_bW}-3|W|^2)\,\label{sgp}
345: \ee
346: near the supersymmetric point, where $D_aW=0$ and $W=W_0$ (the index $a$ 
347: labels the moduli $\phi_a=(\tau,z_i)$). The scalar mass matrix near this 
348: point gets an ${\cal O}(1)$ contribution from the first term, which is 
349: positive definite since the inverse K\"ahler metric has this feature. It 
350: also gets (potentially negative) contributions 
351: \be
352: \sim -e^KK_{a\ol{b}}|W_0|^2\qquad\mbox{and}\qquad\sim -e^K(D_aD_bW)
353: \bar{W}_0
354: \ee
355: from the second term (where we again used the fact that $D_aW$ vanishes in 
356: the vacuum). Thus, in the generic case, all masses should be positive and 
357: ${\cal O}(1)$ if $W_0$ is parametrically small. (This can also be argued
358: by appealing to the known stability of supersymmetric vacua in combination 
359: with the Breitenlohner-Freedman bound~\cite{Breitenlohner:1982bm}: If all 
360: mass squares are ${\cal O}(1)$ and $\Lambda_{\rm AdS}$ is small, all mass 
361: squares must be positive.)
362: 
363: However, in the conifold limit of the one-modulus case analyzed 
364: in~\cite{Denef:2004ze}, this naive expectation was found to be 
365: violated and tachyonic directions (implying the danger of physical 
366: instabilities after uplifting) were found to be generically present. 
367: At the same time, it was argued that this problem will not persist in 
368: models with more than one complex structure modulus. We agree with this
369: expectation and we would like to supply an explicit argument in its favour:
370: 
371: Ignoring the ${\cal O}(1)$ prefactor $e^K$ and using the parametric smallness 
372: of $W_0$, the second-order expression for the supergravity scalar potential, 
373: Eq.~(\ref{sgp}), takes the form 
374: \be
375: V\sim \delta\phi_a\,W_{ab}\,K^{b\ol{c}}\,\ol{W}_{\ol{c}\ol{d}}\,\delta 
376: \ol{\phi}_{\ol{d}}\label{mm}
377: \ee
378: near the vacuum. Here $\delta\phi_a$ is the deviation of $\tau$ (for $a=1$) 
379: or any of the complex structure moduli (for $a>1$) from its vacuum value. 
380: This is, of course, just the familiar rigid-SUSY expression. It is obvious 
381: from Eq.~(\ref{mm}) that a parametrically small eigenvalue of $W_{ab}$ leads 
382: to a light scalar field which, taking into account the full supergravity 
383: expression (including non-vanishing $W_0$) and the uplift, entails the 
384: risk of a tachyonic direction. Indeed, such a small eigenvalue arises in 
385: the one-complex-structure-modulus case near the conifold point: The explicit 
386: form of the superpotential
387: \be
388: W=A(z)+\tau B(z)\,,
389: \ee
390: implies $W_{11}=0$ and the singular behaviour of the integral over the 
391: dual conifold cycle, 
392: \be
393: \int_B\Omega=\frac{z}{2\pi i}\log(z)\,+\,\mbox{holomorphic}\,,\label{lnz}
394: \ee
395: implies $W_{22}\sim 1/z$. The 2$\times$2 matrix $W_{ab}$ then develops a 
396: parametrically small eigenvalue by the usual see-saw mechanism, which 
397: makes the tachyonic direction found in~\cite{Denef:2004ze} possible. 
398: 
399: The situation changes drastically in the case of two or more complex 
400: structure moduli. Let $W_{ab}$ be an $n\times n$ matrix and let $a=n$ 
401: correspond to the conifold modulus $z$ of Eq.~(\ref{lnz}). While it is 
402: still true that $W_{11}=0$ and $W_{nn}$ is parametrically large, this no 
403: longer implies the existence of a small eigenvalue. This can be seen by 
404: considering the characteristic equation
405: \be
406: \mbox{det}(W_{ab}-\lambda\delta_{ab})=0\,.
407: \ee
408: Clearly, the largeness of $W_{nn}$ implies the existence of a large 
409: eigenvalue $\lambda\simeq W_{nn}$. Any further eigenvalue, however, has to 
410: solve the equation 
411: \be
412: \mbox{det}\left(W_{ab}-\lambda\delta_{ab}\big|_{\{a,b=1\ldots n-1\}}\right)
413: =0\,,
414: \ee
415: approximately. The solutions are simply the eigenvalues of an $(n-1)\times 
416: (n-1)$ matrix with vanishing upper-left element, which is otherwise generic. 
417: For $n>1$, neither of these eigenvalues is generically small. Thus, we have 
418: no reason to expect that the problem of a tachyonic direction observed in the 
419: case of a single complex structure modulus will persist. 
420: 
421: Independently of the above, it may also be useful to observe that the 
422: special feature $W_{11}=0$ of the KKLT construction is not generic and can
423: easily be avoided, e.g., by including gaugino condensation on stacks of D3 
424: branes at singularities.\footnote{
425: Related 
426: discussions of the stability of the KKLT construction can be found, e.g., 
427: in~\cite{Brustein:2004xn,Choi:2004sx}.
428: } 
429: 
430: Our main conclusion for the following is that the difficulties 
431: observed in the one-modulus case do not represent an argument against the 
432: existence of many uplifted near-conifold vacua of fluxed multi-modulus 
433: Calabi-Yaus.
434: 
435: 
436: 
437: 
438: 
439: \section{The distribution of throats}
440: We now turn to our main point, which is the interplay between the expected 
441: large number of 3-cycles (and hence of potential conifold singularities) 
442: and the enhancement of the number of vacua in the vicinity of each of those 
443: singularities.
444: 
445: Consider the complex structure moduli space of an orientifold model with 
446: $\sim K$ $\mbox{3-cycles}$, as discussed in the previous section. We assume, 
447: motivated by the example of the quintic and the `Web of Calabi-Yaus', that
448: an ${\cal O}(1)$ fraction of these cycles produce, when they shrink, conifold 
449: singularities. Furthermore, we excise all large complex structure regions, 
450: ending up with a compact moduli space of complex dimension $K/2$. The various 
451: conifold points are described by ${\cal O}(K)$ subspaces of complex 
452: co-dimension one which, in general, intersect each other.
453: 
454: Let us first focus on one of these conifold points (more properly: on one 
455: of the subspaces along which a certain conical singularity persists) and 
456: parameterize the moduli space such that the coordinate $z_i$ characterizes 
457: the shrinking cycle. Making use of the distribution of vacua near a conifold 
458: singularity implied by Eq.~(\ref{enh}), we expect that a randomly chosen 
459: flux vacuum will have probability 
460: \be
461: p_i(|z_i|)\simeq \frac{1}{c_i\log(1/|z_i|)} \label{prob}
462: \ee
463: to be less than $|z_i|$ away from the conifold point under consideration.
464: 
465: The real constant $c_i$ is related to the detailed distribution of vacua 
466: away from the conifold point and to the ambiguities which arise in excluding 
467: the large complex structure regions. To see this, assume for simplicity 
468: that we have excised the region $|z_i|>|z_{i\,,\,max}|$. Clearly, away 
469: from the small-$z_i$ region $p_i$ has some more complicated functional form 
470: (not explicitly known in general) and it has to satisfy the normalization 
471: condition 
472: \be
473: p_i(|z_i|)\to 1\qquad\mbox{for}\qquad |z_i|\to|z_{i\,,\,max}|\,.
474: \ee
475: All that we can infer from Eq.~(\ref{enh}) is that $p_i$ is {\it 
476: proportional} to $1/\log(1/|z_i|)$ at small $z_i$; the normalization is 
477: inextricably linked to the behaviour of $p_i$ at $|z_i|\sim|z_{i\,,\,max}|
478: \sim{\cal O}(1)$. What is worse, $c_i$ is in general a function of the other 
479: complex structure moduli, $c_i=c_i(z_1,\dots,z_{i-1},z_{i+1},\ldots, 
480: z_{K/2})$ (which we ignored in the above), thereby making a detailed analysis 
481: of the full probability distribution highly non-trivial. Thus, all that we 
482: can do at the moment is to assume that the various $c_i$ do not vary too 
483: rapidly and are not parametrically large or small (for which there is no 
484: obvious reason). We will parameterize our ignorance assigning a universal 
485: unknown value of the order of one to all these coefficients, $c_i=c$. 
486: 
487: Let us now recall that, if $z_i$ is stabilized near zero, a strongly warped 
488: region or throat with a hierarchy of mass scales 
489: \be
490: h_i \sim |z_i|^{-1/3}
491: \ee
492: between the Klebanov-Strassler region (IR end) and the Calabi-Yau region 
493: (UV end) develops~\cite{Giddings:2001yu} (see also~\cite{DeWolfe:2002nn}). 
494: We conclude from the above that the probability for finding a throat with a 
495: hierarchy larger than $h_i$ is
496: \be
497: p_i(h_i)\simeq\frac{1}{3c_i\log\,h_i}\,.
498: \ee
499: Intuitively, this characterizes the probability for being, in the given 
500: moduli space, within a slice of a certain thickness that surrounds the 
501: hypersurface defined by $z_i=0$. 
502: 
503: Assuming that these probabilities are uncorrelated for the various 
504: conifold hypersurfaces (i.e. for the various $z_i,\, i=1\ldots K$), we can 
505: estimate the probability for finding precisely $n$ throats with hierarchy 
506: larger than $h_*$. It is given by the probability for being inside $n$ of the 
507: $K$ slices and outside the remaining $K-n$ slices, multiplied by a 
508: combinatorial factor for choosing inside which slices to be:
509: \be
510: p(n,h>h_*|K)\sim \binom{K}{n}\,p^n(1-p)^{K-n}\qquad\mbox{with}\qquad
511: p\equiv\frac{1}{3c\log h_*}\,.
512: \ee
513: The fact that this `multi-throat probability' is given simply by a 
514: binomial distribution with parameters $K$ and $p$ represents one of our main 
515: results (or, given the various assumptions above, our main conjecture). 
516: Many interesting and potentially phenomenologically relevant questions 
517: can now be addressed.
518: 
519: For example, given a certain hierarchy factor $h_*$, we can inquire about the 
520: expected number of throats with a larger hierarchy. It is given by the 
521: well-known mean of the binomial distribution,
522: \be
523: \ol{n}(h>h_*|K)=\sum_{n=0}^K n\,\binom{K}{n}\,p^n(1-p)^{K-n}=Kp=\frac{K}
524: {3c\log h_*}\,.\label{nbar}
525: \ee
526: The crucial but certainly not unexpected point here is that $\bar{n}$ goes
527: to zero very slowly as $h_*$ grows. The variance of $n$, again a familiar 
528: result, is
529: \be
530: \mbox{var}(n)=\sum_{n=0}^K(n-\ol{n})^2\,\binom{K}{n}\,p^n(1-p)^{K-n}=
531: Kp(1-p)\,,
532: \ee
533: which is very close to $\ol{n}$ for $p\ll 1$. Thus, the expected number of 
534: throats is
535: \be
536: \ol{n}\pm\sqrt{\ol{n}}\,,
537: \ee
538: with $\ol{n}$ as given in Eq.~(\ref{nbar}).
539: 
540: For a given $K$ the probability that at least one throat has hierarchy 
541: exceeding some specified $h_*$ is given by
542: \be
543: P(h>h_*|K) = \left(1-\frac{1}{3c\log h_*}\right)^K \left[ \left(1+\frac{1}
544: {3 c \log h_* -1}\right)^K -1\right]\,.\label{atleast}
545: \ee
546: Figure~1 shows this function against $\log h_*$ for $c=1$ and $K$
547: taking the values $60$ and $200$. It is noteworthy that there is a very 
548: slow decrease of the probability with throat length,
549: and that at 50\% likelihood there exist throats of hierarchy greater than
550: $\exp(28)\sim 10^{12}$, and $\exp(95)\sim 10^{41}$ as
551: $K$ varies from 60 to 200. Not surprisingly, Eq.~(\ref{atleast}) coincides 
552: with Eq.~(\ref{nbar}) if $K/3c\log(h_*)\ll 1$.
553:   
554: \begin{figure}
555: \begin{center}
556: \includegraphics[width=10cm]{hierarchy.eps}
557: \caption{Probability that at least one throat has hierarchy $h>h_*$ as a 
558: function of $\log h_*$. Reading from bottom to top the curves correspond to 
559: the choices $K=60$ and 200 (both taking $c=1$).}
560: \label{pot}
561: \end{center}
562: \end{figure}
563: 
564: Another interesting quantity is the hierarchy $h_1$ of the longest expected
565: throat. A simply estimate of this quantity is provided by solving 
566: \be
567: \ol{n}(h>h_1|K)\sim 1
568: \ee
569: for $h_1$. The result is
570: \be
571: \log h_1\sim\frac{K}{3c}\,.\label{h1}
572: \ee
573: Alternatively, we can ask for which $h_1$ the one-throat-probability is 
574: maximized,
575: \be
576: \frac{d}{d\,h_1}\,p(1,h>h_1|K)=0\,.
577: \ee
578: The result is consistent with Eq.~(\ref{h1}). Yet another way to state 
579: the same problem is to ask, at fixed $h_1$, for the value of $K$ which gives 
580: the maximal value for $p(1,h>h_1|K)$. Again, the resulting relation of $h_1$ 
581: and $K$ is approximately that of Eq.~(\ref{h1}). 
582: 
583: Furthermore, a very simple but important quantity is the probability of having 
584: no throat with a hierarchy larger than $h_*$,
585: \be
586: p(0,h>h_*|K)\simeq (1-p)^K \simeq \exp\left(-\frac{K}{3c\log h_*}\right)
587: \,.\label{p0}
588: \ee
589: As expected, this is a very small number for large $K$ and not too large
590: hierarchies. We consider this together with the expected number of throats, 
591: Eq.~(\ref{nbar}), the expected hierarchy of the longest throat, 
592: Eq.~(\ref{h1}), and the one-throat-probability of Fig.~\ref{pot} to be the 
593: main results of this section. 
594: 
595: Finally, using Eq.~(\ref{atleast}) it might also be possible to gain
596: information on $K$ independent of the fine-tuning of the cosmological constant
597: by conditioning on the existence of an electro-weak throat with IR 
598: scale $\sim$TeV. Using Bayes' theorem the conditional probability 
599: distribution for $K$ given that there exists at least one throat with $h\geq 
600: h_{EW}$ is
601: \be
602: P(K|n(h_{EW})\geq 1) = \frac{P(h>h_{EW}|K) P(K)}{\Sigma_{K'=1}^{K_{max}} 
603: P(h>h_{EW}|K') P(K')} .\label{ewprior}
604: \ee
605: If we conservatively assume a flat prior distribution for $K$, $P(K) = 
606: 1/K_{max}$ and take as an illustrative example $K_{max}=200$ and $c=1$, then 
607: a numerical evaluation of Eq.~(\ref{ewprior}) leads to an {\it a posteriori} 
608: mean ${\bar K} \sim 124$. 
609: 
610: 
611: 
612: 
613: \section{Possible phenomenological implications}
614: To discuss possible phenomenological implications, we have to quantify the 
615: expected hierarchies $h$, which depend crucially on the number of cycles $K$. 
616: Since, at the fundamental level, we are ignorant about $K$ and, moreover, $K$
617: appears in the combination $K/3c$ (with an unknown ${\cal O}(1)$-constant 
618: $c$), we take the following pragmatic approach:
619: 
620: We consider two scenarios, one conservative and one more favourable: In the 
621: conservative case, we choose $K=60$ (roughly the minimal value consistent 
622: with fine-tuning $\Lambda$) and 
623: $c=3$, such that the relevant combination of these two parameters takes the 
624: low value $K/3c\simeq 7$. In the favourable case, we choose $K=200$ 
625: (consistent with typical Calabi-Yau values, maybe somewhat at the high side, 
626: but not extreme). Together with $c=1/3$ this gives the high value 
627: $K/3c\simeq 200$. 
628: 
629: In the conservative case, Eq.~(\ref{h1}) implies that the longest throat 
630: typically has a hierarchy $\sim 10^3$. This clearly also means that,
631: specifying a minimal hierarchy $10^3$, we expect about one throat with 
632: a hierarchy above that value. We can also infer that we have to expect about 
633: 3 throats with hierarchy 10 or larger. Even though these numbers are not 
634: very impressive, they clearly imply that dynamically generated scales 
635: of $\sim 10^{-3}M_P$ are natural in the present branch of the landscape. 
636: The above short throats can play an important role in inflation or 
637: simply to ensure a small (and hence perturbatively controlled) anti-D3-brane 
638: uplift. Thus, even though no spectacular low-energy effects can be predicted 
639: in this conservative setting, moderate throats are indeed `ubiquitous'. 
640: 
641: What is maybe more impressive is the small statistical price that one pays 
642: for having a moderately long throat. For example, the expectation value for 
643: the number of throats with hierarchy above $10^6$ is $0.5$. In other words, 
644: low scale SUSY in the KKLT setting (see e.g.~\cite{Choi:2005ge}) is perfectly 
645: plausible and does not require any extra fine-tuning. Even more, demanding a 
646: hierarchy of $10^{13}$ or higher, one still finds an expectation value 
647: for the throat number of approximately $0.23$. In other words, generating 
648: the electroweak hierarchy is also very plausible since about 1 in 4
649: vacua have a sufficiently long throat. However, we can clearly not claim 
650: that throats of this length are unavoidable. 
651: 
652: We now turn to the case $K/3c=200$, where things look very different 
653: indeed. The longest expected throat produces a huge hierarchy $\sim 10^{80}$. 
654: Thus, we expect almost conformal field theories with very low IR cutoff (which 
655: are presumably only gravitationally coupled to standard model matter) to 
656: be abundant. Specifically, not having a throat with hierarchy $10^{29}$ or 
657: larger (corresponding to an IR scale of meV) has probability of about 5\%
658: (cf. Eq.~(\ref{p0})). In other words, hidden sectors with dynamical scales 
659: $\sim$meV or below are a {\it prediction} of this setting. Clearly, the above 
660: can have very important cosmological implications as far as dark matter or 
661: dark radiation are concerned. Just to give one more numerical implication 
662: of the formulae of the last section: The expected number of throats with a
663: hierarchy larger than $M_P/M_{EW}\sim 10^{14}$ is $\bar{n}\sim 6$. Thus, 
664: several electroweak-scale hidden sectors are a natural occurrence. The 
665: phenomenological and cosmological implications of this scenario clearly 
666: depend very strongly on whether `we' are in the throat or on the Calabi-Yau, 
667: where inflation took place and how strongly throat sectors are coupled 
668: to each other and to light fields localized in the UV. 
669: Away from the cosmological context there are two outstanding possibilities
670: for signatures of long throats with IR scale at or below the weak scale which
671: have been partially investigated: invisible Higgs decays to hidden sector 
672: particles\cite{Patt:2006fw}, and kinetic mixing of hypercharge with 
673: hidden-sector U(1)'s\cite{Kumar:2006gm}.  All we can say at 
674: present is that the various scenarios of this type studied in the 
675: literature appear to be everything else but exotic. 
676: 
677: 
678: \section{Conclusions}
679: Based on a number of assumptions, we have quantified the expectation that 
680: throats are common in the type IIB landscape. The crucial starting point 
681: is the large number of 3-cycles which the compact space is expected to have. 
682: This can be quantified in two ways: conservatively, by taking the minimal
683: number which allows for the fine-tuning of $\Lambda$, or more optimistically, 
684: by taking a number which is typical for the more complicated Calabi-Yau 
685: manifolds. Given this large number of cycles (all of which generically 
686: carry a certain discrete flux number), one has to expect that by pure 
687: chance the flux on some of these cycles will be relatively small. Those
688: cycles are stabilized at small size, which generically leads to the 
689: development of a throat and a large hierarchy of scales. We have made this 
690: last argument more precise on the basis of the known behaviour of the 
691: density of flux vacua near conifold points. 
692: 
693: Our main technical results are simple formulae for the expectation value of
694: the number of throats with a certain hierarchy and for the probability of 
695: having no throat with a hierarchy larger than some given value. The
696: numerical predictions depend on the uncertain total number of 3-cycles 
697: mentioned above and on the details of the flux distribution away from the 
698: conifold points. Even with conservative assumptions about both of these 
699: unknown quantities, short throats (with hierarchies $\sim 10^3$) are 
700: generically expected while longer throats (with electroweak hierarchy) 
701: are at least not uncommon. Taking optimistic values for the unknown 
702: input data, we find that extreme hierarchies $\sim 10^{80}$ are expected and 
703: throats with electroweak hierarchy represent a firm statistical prediction. 
704: 
705: While these findings confirm the claim of our title that throats are 
706: ubiquitous in the type IIB landscape, the typical length of those throats
707: is quite uncertain at present. In this respect the main open questions are 
708: how complex a Calabi-Yau we should be looking for and a quantitative 
709: understanding of the `bulk' of the high-dimensional moduli space of such
710: manifolds. Furthermore, a better understanding of the role played by the 
711: large complex structure regions (which we have ignored in this analysis)
712: is highly desirable. 
713: 
714: Finally, even though many important questions remain unanswered, we consider 
715: one conclusion as relatively firm: Throats are common in the presently best
716: understood part of the string-theory landscape and should thus be taken 
717: very seriously both in string-theoretic and phenomenological model building. 
718: Given the very general setting we have been working with and the small
719: number of assumptions that we had to make, we are optimistic that 
720: throats will become one of the most firm and concrete predictions of the 
721: type IIB landscape. We would like to view this as strong support for 
722: phenomenological research in 5d Randall-Sundrum-like model, put within the 
723: more specific limits of their type IIB realization~\cite{Gherghetta:2006yq}. 
724: At the same time one should, however, keep in mind that, if the `favourable'
725: scenario of many very long throats is confirmed and cosmological problems 
726: with the various light fields are established, this whole line of thinking may 
727: turn into a serious argument against the type IIB landscape. 
728: 
729: \noindent
730: {\bf Acknowledgements}:\hspace*{.5cm}We would like to thank Savas Dimopoulos 
731: and Maximilian Kreuzer for helpful discussions.  We thank the Galileo Galilei
732: Institute for Theoretical Physics for hospitality and the INFN for partial
733: support during the completion of this work.
734: 
735: %\vspace*{-.5cm}
736: \begin{thebibliography}{99}
737: 
738: \bibitem{Giddings:2001yu}
739:   S.~B.~Giddings, S.~Kachru and J.~Polchinski,
740:   ``Hierarchies from fluxes in string compactifications,''
741:   Phys.\ Rev.\ D {\bf 66} (2002) 106006 [arXiv:hep-th/0105097].
742:   %%CITATION = HEP-TH 0105097;%%
743: 
744: \bibitem{earlywarp}
745:   S.~Sethi, C.~Vafa and E.~Witten,
746:   ``Constraints on low-dimensional string compactifications,''
747:   Nucl.\ Phys.\ B {\bf 480} (1996) 213
748:   [arXiv:hep-th/9606122];\\
749:   %%CITATION = HEP-TH 9606122;%%
750:   K.~Dasgupta, G.~Rajesh and S.~Sethi,
751:   ``M theory, orientifolds and G-flux,''
752:   JHEP {\bf 9908} (1999) 023
753:   [arXiv:hep-th/9908088];\\
754:   %%CITATION = HEP-TH 9908088;%%
755:   B.~R.~Greene, K.~Schalm and G.~Shiu,
756:   ``Warped compactifications in M and F theory,''
757:   Nucl.\ Phys.\ B {\bf 584}, 480 (2000)
758:   [arXiv:hep-th/0004103].
759:   %%CITATION = HEP-TH 0004103;%%
760: 
761: \bibitem{Verlinde:1999fy}
762:   H.~L.~Verlinde, ``Holography and compactification,''
763:   Nucl.\ Phys.\ B {\bf 580}, 264 (2000) [arXiv:hep-th/9906182];\\
764:   %%CITATION = HEP-TH 9906182;%%
765: %\bibitem{Chan:2000ms}
766:   C.~S.~Chan, P.~L.~Paul and H.~L.~Verlinde, ``A note on warped string
767:   compactification,'' Nucl.\ Phys.\ B {\bf 581} (2000) 156
768:   [arXiv:hep-th/0003236].
769:   %%CITATION = HEP-TH 0003236;%%
770:   
771: \bibitem{Grana:2005jc}
772:   M.~Grana, ``Flux compactifications in string theory: A comprehensive 
773:   review,'' Phys.\ Rept.\  {\bf 423} (2006) 91 [arXiv:hep-th/0509003].
774:   %%CITATION = HEP-TH 0509003;%%
775: 
776: \bibitem{Kachru:2003aw}
777:   S.~Kachru, R.~Kallosh, A.~Linde and S.~P.~Trivedi, ``De Sitter vacua in 
778:   string theory,'' Phys.\ Rev.\ D {\bf 68} (2003) 046005
779:   [arXiv:hep-th/0301240].
780:   %%CITATION = HEP-TH 0301240;%%
781: 
782: \bibitem{landscape}
783:   R.~Bousso and J.~Polchinski,
784:    ``Quantization of four-form fluxes and dynamical neutralization of the
785:   cosmological constant,''
786:   JHEP {\bf 0006} (2000) 006
787:   [arXiv:hep-th/0004134];\\
788:   %%CITATION = HEP-TH 0004134;%%
789: J.~L.~Feng, J.~March-Russell, S.~Sethi and F.~Wilczek,
790:   ``Saltatory relaxation of the cosmological constant,''
791:   Nucl.\ Phys.\ B {\bf 602} (2001) 307
792:   [arXiv:hep-th/0005276].
793:   %%CITATION = HEP-TH 0005276;%%
794: 
795: 
796: 
797:   
798: \bibitem{Douglas:2003um}
799:   M.~R.~Douglas, ``The statistics of string / M theory vacua,''
800:   JHEP {\bf 0305} (2003) 046 [arXiv:hep-th/0303194];\\
801:   %%CITATION = HEP-TH 0303194;%%
802: %\bibitem{Ashok:2003gk}
803:   S.~Ashok and M.~R.~Douglas, ``Counting flux vacua,''
804:   JHEP {\bf 0401} (2004) 060 [arXiv:hep-th/0307049].
805:   %%CITATION = HEP-TH 0307049;%%
806: 
807: \bibitem{Denef:2004ze}
808:   F.~Denef and M.~R.~Douglas, ``Distributions of flux vacua,''
809:   JHEP {\bf 0405}, 072 (2004) [arXiv:hep-th/0404116]. 
810:   %%CITATION = HEP-TH 0404116;%%
811: 
812: \bibitem{Denef:2004cf}
813:   F.~Denef and M.~R.~Douglas, ``Distributions of nonsupersymmetric flux 
814:   vacua,'' JHEP {\bf 0503} (2005) 061 [arXiv:hep-th/0411183].
815:   %%CITATION = HEP-TH 0411183;%%
816: 
817: \bibitem{Balasubramanian:2004uy}
818:   V.~Balasubramanian and P.~Berglund, ``Stringy corrections to Kaehler 
819:   potentials, SUSY breaking, and the cosmological constant problem,''
820:   JHEP {\bf 0411} (2004) 085 [arXiv:hep-th/0408054];\\
821:   %%CITATION = HEP-TH 0408054;%%
822: %\bibitem{Bobkov:2004cy}
823:   K.~Bobkov, ``Volume stabilization via alpha' corrections in type IIB theory
824:   with fluxes,'' JHEP {\bf 0505} (2005) 010 [arXiv:hep-th/0412239].
825:   %%CITATION = HEP-TH 0412239;%%
826: 
827: \bibitem{Balasubramanian:2005zx}
828:   V.~Balasubramanian, P.~Berglund, J.~P.~Conlon and F.~Quevedo, ``Systematics
829:   of moduli stabilisation in Calabi-Yau flux compactifications,'' JHEP {\bf
830:   0503} (2005) 007 [arXiv:hep-th/0502058];\\
831:   %%CITATION = HEP-TH 0502058;%%
832: %\bibitem{Conlon:2005ki}
833:   J.~P.~Conlon, F.~Quevedo and K.~Suruliz,
834:   ``Large-volume flux compactifications: Moduli spectrum and D3/D7 soft
835:   supersymmetry breaking,'' JHEP {\bf 0508} (2005) 007 [arXiv:hep-th/0505076].
836:   %%CITATION = HEP-TH 0505076;%%
837: 
838: \bibitem{Saltman:2004sn}
839:   A.~Saltman and E.~Silverstein, ``The scaling of the no-scale potential and
840:   de Sitter model building,'' JHEP {\bf 0411} (2004) 066
841:   [arXiv:hep-th/0402135].
842:   %%CITATION = HEP-TH 0402135;%%
843: 
844: \bibitem{vonGersdorff:2005bf}
845:   G.~von Gersdorff and A.~Hebecker, ``Kaehler corrections for the volume
846:   modulus of flux compactifications,'' Phys.\ Lett.\ B {\bf 624} (2005) 270
847:   [arXiv:hep-th/0507131];\\
848:   %%CITATION = HEP-TH 0507131;%%
849: %\bibitem{Berg:2005ja}
850:   M.~Berg, M.~Haack and B.~Kors, ``String loop corrections to Kaehler
851:   potentials in orientifolds,'' JHEP {\bf 0511} (2005) 030
852:   [arXiv:hep-th/0508043]
853:   %%CITATION = HEP-TH 0508043;%%
854: and
855: %\bibitem{Berg:2005yu} M.~Berg, M.~Haack and B.~Kors,
856:   ``On volume stabilization by quantum corrections,'' Phys.\ Rev.\ Lett.\ 
857:   {\bf 96} (2006) 021601 [arXiv:hep-th/0508171];\\
858:   %%CITATION = HEP-TH 0508171;%%
859: %\bibitem{Parameswaran:2006jh}
860:   S.~L.~Parameswaran and A.~Westphal, ``de Sitter string vacua from 
861:   perturbative Kaehler corrections and consistent D-terms,''
862:   arXiv:hep-th/0602253.
863:   %%CITATION = HEP-TH 0602253;%%
864: 
865: \bibitem{Candelas:1989js}
866:   P.~Candelas and X.~C.~de la Ossa,
867:   ``Comments On Conifolds,''
868:   Nucl.\ Phys.\ B {\bf 342} (1990) 246.
869:   %%CITATION = NUPHA,B342,246;%%
870: 
871: \bibitem{Klebanov:1998hh}
872:   I.~R.~Klebanov and E.~Witten, ``Superconformal field theory on threebranes
873:   at a Calabi-Yau  singularity,'' Nucl.\ Phys.\ B {\bf 536} (1998) 199
874:   [arXiv:hep-th/9807080];\\
875:   %%CITATION = HEP-TH 9807080;%%
876: %\bibitem{Klebanov:1999rd}
877:   I.~R.~Klebanov and N.~A.~Nekrasov, ``Gravity duals of fractional branes and 
878:   logarithmic RG flow,'' Nucl.\ Phys.\ B {\bf 574} (2000) 263
879:   [arXiv:hep-th/9911096];\\
880:   %%CITATION = HEP-TH 9911096;%%
881: %\bibitem{Klebanov:2000nc}
882:   I.~R.~Klebanov and A.~A.~Tseytlin, ``Gravity duals of supersymmetric SU(N)
883:   x SU(N+M) gauge theories,'' Nucl.\ Phys.\ B {\bf 578} (2000) 123
884:   [arXiv:hep-th/0002159].
885:   %%CITATION = HEP-TH 0002159;%%
886: 
887: \bibitem{Klebanov:2000hb}
888:   I.~R.~Klebanov and M.~J.~Strassler, ``Supergravity and a confining gauge 
889:   theory: Duality cascades and $\chi$SB-resolution of naked singularities,''
890:   JHEP {\bf 0008} (2000) 052 [arXiv:hep-th/0007191].
891:   %%CITATION = HEP-TH 0007191;%%
892: 
893: \bibitem{Randall:1999ee}
894:   L.~Randall and R.~Sundrum, ``A large mass hierarchy from a small extra 
895:   dimension,'' Phys.\ Rev.\ Lett.\  {\bf 83} (1999) 3370
896:   [arXiv:hep-ph/9905221] and 
897:   %%CITATION = HEP-PH 9905221;%%
898: %\bibitem{Randall:1999vf} L.~Randall and R.~Sundrum, 
899:   ``An alternative to compactification,'' Phys.\ Rev.\ Lett.\  {\bf 83} 
900:   (1999) 4690 [arXiv:hep-th/9906064].
901:   %%CITATION = HEP-TH 9906064;%%
902: 
903: 
904: 
905: \bibitem{Goldberger:1999uk}
906:   W.~D.~Goldberger and M.~B.~Wise, ``Modulus stabilization with bulk fields,''
907:   Phys.\ Rev.\ Lett.\  {\bf 83} (1999) 4922 [arXiv:hep-ph/9907447].
908:   %%CITATION = HEP-PH 9907447;%%
909: 
910: \bibitem{smallnumbers}
911:   S.~Dimopoulos, S.~Kachru, N.~Kaloper, A.~E.~Lawrence and E.~Silverstein,
912:   ``Small numbers from tunneling between brane throats,''
913:   Phys.\ Rev.\ D {\bf 64} (2001) 121702
914:   [arXiv:hep-th/0104239]; ``Generating small numbers by tunneling in 
915:   multi-throat  compactifications,
916:   Int.\ J.\ Mod.\ Phys.\ A {\bf 19} (2004) 2657
917:   [arXiv:hep-th/0106128].
918:   %%CITATION = HEP-TH 0106128;%%
919:   
920: 
921: 
922: \bibitem{Barnaby:2004gg}
923:   N.~Barnaby, C.~P.~Burgess and J.~M.~Cline, ``Warped reheating in 
924:   brane-antibrane inflation,'' JCAP {\bf 0504}, 007 (2005)
925:   [arXiv:hep-th/0412040];\\
926:   %%CITATION = HEP-TH 0412040;%%
927: %\bibitem{Kofman:2005yz}
928:   L.~Kofman and P.~Yi, ``Reheating the universe after string theory
929:   inflation,'' Phys.\ Rev.\ D {\bf 72} (2005) 106001
930:   [arXiv:hep-th/0507257];\\
931:   %%CITATION = HEP-TH 0507257;%%
932: %\bibitem{Frey:2005jk}
933:   A.~R.~Frey, A.~Mazumdar and R.~Myers, ``Stringy effects during inflation
934:   and reheating,'' Phys.\ Rev.\ D {\bf 73} (2006) 026003
935:   [arXiv:hep-th/0508139];\\
936:   %%CITATION = HEP-TH 0508139;%%
937: %\bibitem{Chialva:2005zy}
938:   D.~Chialva, G.~Shiu and B.~Underwood,
939:   ``Warped reheating in multi-throat brane inflation,''
940:   JHEP {\bf 0601} (2006) 014
941:   [arXiv:hep-th/0508229];\\
942:   %%CITATION = HEP-TH 0508229;%%
943: %\bibitem{Firouzjahi:2005qs}
944:   H.~Firouzjahi and S.~H.~Tye,
945:   ``The shape of gravity in a warped deformed conifold,''
946:   JHEP {\bf 0601} (2006) 136
947:   [arXiv:hep-th/0512076].
948:   %%CITATION = HEP-TH 0512076;%%
949: %\bibitem{Chen:2006ni}
950:   X.~Chen and S.~H.~Tye, ``Heating in brane inflation and hidden dark
951:   matter,'' arXiv:hep-th/0602136.
952:   %%CITATION = HEP-TH 0602136;%%
953: 
954: \bibitem{Cascales:2003wn}
955:   J.~F.~G.~Cascales, M.~P.~Garcia del Moral, F.~Quevedo and A.~M.~Uranga,
956:   ``Realistic D-brane models on warped throats: Fluxes, hierarchies and
957:   moduli stabilization,'' JHEP {\bf 0402} (2004) 031 [arXiv:hep-th/0312051];\\
958: %\bibitem{Cacciapaglia:2006tg}
959:   G.~Cacciapaglia, C.~Csaki, C.~Grojean and J.~Terning,
960:    ``Field theory on multi-throat backgrounds,'' arXiv:hep-ph/0604218.
961:   %%CITATION = HEP-PH 0604218;%%
962: 
963: \bibitem{Giryavets:2004zr}
964:   A.~Giryavets, S.~Kachru and P.~K.~Tripathy, ``On the taxonomy of flux 
965:   vacua,'' JHEP {\bf 0408} (2004) 002 [arXiv:hep-th/0404243].
966:   %%CITATION = HEP-TH 0404243;%%
967: 
968: \bibitem{Conlon:2004ds}
969:   J.~P.~Conlon and F.~Quevedo, ``On the explicit construction and statistics
970:   of Calabi-Yau flux vacua,'' JHEP {\bf 0410} (2004) 039
971:   [arXiv:hep-th/0409215].
972:   %%CITATION = HEP-TH 0409215;%%
973: 
974: \bibitem{Eguchi:2005eh}
975:   T.~Eguchi and Y.~Tachikawa,
976:   ``Distribution of flux vacua around singular points in Calabi-Yau moduli
977:   space,'' JHEP {\bf 0601} (2006) 100 [arXiv:hep-th/0510061].
978:   %%CITATION = HEP-TH 0510061;%%
979: 
980: \bibitem{Kachru:2002gs}
981:   S.~Kachru, J.~Pearson and H.~L.~Verlinde, ``Brane/flux annihilation and
982:   the string dual of a non-supersymmetric  field theory,''
983:   JHEP {\bf 0206} (2002) 021 [arXiv:hep-th/0112197];\\
984:   %%CITATION = HEP-TH 0112197;%%
985: %\bibitem{Frey:2003dm}
986:   A.~R.~Frey, M.~Lippert and B.~Williams, ``The fall of stringy de Sitter,''
987:   Phys.\ Rev.\ D {\bf 68} (2003) 046008 [arXiv:hep-th/0305018].
988:   %%CITATION = HEP-TH 0305018;%%
989: 
990: \bibitem{Klemm:1996ts}
991:   A.~Klemm, B.~Lian, S.~S.~Roan and S.~T.~Yau, ``Calabi-Yau fourfolds for 
992:   M- and F-theory compactifications,'' Nucl.\ Phys.\ B {\bf 518} (1998) 515
993:   [arXiv:hep-th/9701023].
994:   %%CITATION = HEP-TH 9701023;%%
995: 
996: \bibitem{Candelas:1994hw}
997:   P.~Candelas, A.~Font, S.~Katz and D.~R.~Morrison,
998:    ``Mirror symmetry for two parameter models. 2,''
999:      Nucl.\ Phys.\ B {\bf 429}, 626 (1994)
1000:   [arXiv:hep-th/9403187].
1001:   %%CITATION = HEP-TH 9403187;%%
1002:   
1003: \bibitem{Avram:1996pj}
1004:   A.~C.~Avram, M.~Kreuzer, M.~Mandelberg and H.~Skarke, ``Searching for K3 
1005:   fibrations,'' Nucl.\ Phys.\ B {\bf 494} (1997) 567 [arXiv:hep-th/9610154].
1006:   %%CITATION = HEP-TH 9610154;%%
1007: 
1008: \bibitem{Denef:2004dm}
1009:   F.~Denef, M.~R.~Douglas and B.~Florea,
1010:    ``Building a better racetrack,''
1011:   JHEP {\bf 0406}, 034 (2004)
1012:   [arXiv:hep-th/0404257].
1013:   %%CITATION = HEP-TH 0404257;%%
1014: 
1015: \bibitem{Lefschetz} S.~Lefschetz, ``L'Analysis Situs et la G\'{e}ometri\'{e} 
1016:   Alg\'{e}brique'', Gauthier-Villars, Paris, 1924; reprinted in ``Selected 
1017:   Papers'', Chelsea, New York, 1971, pp. 283-439.
1018: 
1019: \bibitem{Candelas:1989ug}
1020:   P.~Candelas, P.~S.~Green and T.~Hubsch, ``Rolling Among Calabi-Yau Vacua,''
1021:   Nucl.\ Phys.\ B {\bf 330}, 49 (1990);\\
1022:   %%CITATION = NUPHA,B330,49;%%
1023: %\bibitem{Greene:1995hu}
1024:   B.~R.~Greene, D.~R.~Morrison and A.~Strominger, ``Black hole condensation
1025:   and the unification of string vacua,'' Nucl.\ Phys.\ B {\bf 451} (1995)
1026:   109 [arXiv:hep-th/9504145].
1027:   %%CITATION = HEP-TH 9504145;%%
1028:   
1029: 
1030: \bibitem{Curio:2000sc}  
1031: For a nice discussion of Calabi-Yau singularities, see for example,
1032:   G.~Curio, A.~Klemm, D.~Lust and S.~Theisen,
1033:    ``On the vacuum structure of type II string compactifications on  Calabi-Yau
1034: spaces with H-fluxes,''
1035:   Nucl.\ Phys.\ B {\bf 609} (2001) 3
1036:   [arXiv:hep-th/0012213].
1037:   %%CITATION = HEP-TH 0012213;%%
1038: 
1039: 
1040: \bibitem{Breitenlohner:1982bm}
1041:   P.~Breitenlohner and D.~Z.~Freedman, ``Positive Energy In Anti-De Sitter
1042:   Backgrounds And Gauged Extended Supergravity,'' Phys.\ Lett.\ B {\bf 115}
1043:   (1982) 197.
1044:   %%CITATION = PHLTA,B115,197;%%
1045: 
1046: \bibitem{Brustein:2004xn}
1047:   R.~Brustein and S.~P.~de Alwis, ``Moduli potentials in string 
1048:   compactifications with fluxes: Mapping the discretuum,''
1049:   Phys.\ Rev.\ D {\bf 69} (2004) 126006 [arXiv:hep-th/0402088].
1050:   %%CITATION = HEP-TH 0402088;%%
1051: 
1052: \bibitem{Choi:2004sx}
1053:   K.~Choi, A.~Falkowski, H.~P.~Nilles, M.~Olechowski and S.~Pokorski,
1054:   ``Stability of flux compactifications and the pattern of supersymmetry
1055:   breaking,'' JHEP {\bf 0411} (2004) 076 [arXiv:hep-th/0411066];\\
1056:   %%CITATION = HEP-TH 0411066;%%
1057: %\bibitem{deAlwis:2005tf}
1058:   S.~P.~de Alwis, ``Effective potentials for light moduli,''
1059:   Phys.\ Lett.\ B {\bf 626} (2005) 223 [arXiv:hep-th/0506266]
1060:   %%CITATION = HEP-TH 0506266;%%
1061: and 
1062: %\bibitem{deAlwis:2006sz}
1063:   ``Potentials for light moduli in supergravity and string theory,''
1064:   arXiv:hep-th/0602182;\\
1065:   %%CITATION = HEP-TH 0602182;%%
1066:   G.~Curio and V.~Spillner,
1067:   ``On the modified KKLT procedure: A case study for the P(11169)(18) model,''
1068:   arXiv:hep-th/0606047;\\
1069:   %%CITATION = HEP-TH 0606047;%%
1070:   H.~Abe, T.~Higaki and T.~Kobayashi,
1071:   ``Remark on integrating out heavy moduli in flux compactification,''
1072:   arXiv:hep-th/0606095.
1073:   %%CITATION = HEP-TH 0606095;%%
1074: 
1075: 
1076: \bibitem{DeWolfe:2002nn}
1077:   O.~DeWolfe and S.~B.~Giddings, ``Scales and hierarchies in warped 
1078:   compactifications and brane worlds,'' Phys.\ Rev.\ D {\bf 67} (2003) 066008
1079:   [arXiv:hep-th/0208123];\\
1080:   %%CITATION = HEP-TH 0208123;%%
1081: %\bibitem{Giddings:2005ff}
1082:   S.~B.~Giddings and A.~Maharana, ``Dynamics of warped compactifications and
1083:   the shape of the warped landscape,'' Phys.\ Rev.\ D {\bf 73} (2006) 126003
1084:   [arXiv:hep-th/0507158].
1085:   %%CITATION = HEP-TH 0507158;%%
1086: 
1087: \bibitem{Choi:2005ge}
1088:   K.~Choi, A.~Falkowski, H.~P.~Nilles and M.~Olechowski,
1089:    ``Soft supersymmetry breaking in KKLT flux compactification,''
1090:   Nucl.\ Phys.\ B {\bf 718} (2005) 113 [arXiv:hep-th/0503216].
1091:   %%CITATION = HEP-TH 0503216;%%
1092: 
1093: \bibitem{Patt:2006fw}
1094:   B.~Patt and F.~Wilczek, ``Higgs-field portal into hidden sectors,''
1095:   arXiv:hep-ph/0605188.
1096:   %%CITATION = HEP-PH 0605188;%%
1097:   
1098: \bibitem{Kumar:2006gm}
1099:   J.~Kumar and J.~D.~Wells, ``LHC and ILC probes of hidden-sector gauge
1100:   bosons,'' arXiv:hep-ph/0606183.
1101:   %%CITATION = HEP-PH 0606183;%%
1102: 
1103: \bibitem{Gherghetta:2006yq}
1104:   T.~Gherghetta and J.~Giedt, ``Bulk fields in AdS(5) from probe D7
1105:   branes,'' arXiv:hep-th/0605212;\\
1106:   %%CITATION = HEP-TH 0605212;%%
1107: %\bibitem{Brummer:2005sh}
1108:   F.~Br\"ummer, A.~Hebecker and E.~Trincherini, ``The throat as a
1109:   Randall-Sund\-rum model with Goldberger-Wise stabilization,'' Nucl.\ Phys.\
1110:   B {\bf 738} (2006) 283 [arXiv:hep-th/0510113];\\
1111:   %%CITATION = HEP-TH 0510113;%%
1112: %\bibitem{Brummer:2006dg}
1113:   F.~Br\"ummer, A.~Hebecker and M.~Trapletti, ``SUSY breaking mediation by
1114:   throat fields,'' arXiv:hep-th/0605232.
1115:   %%CITATION = HEP-TH 0605232;%%
1116:   
1117: \end{thebibliography}
1118: \end{document}
1119: