Relatively hyperbolic metric bundles and Cannon-Thurston map

Let (X,B,\pi) be a metric (graph) bundle satisfying the following.

1. B is a hyperbolic metric space.

2. Each fiber F_b, b\in B, is strongly hyperbolic relative to a collection of uniformly mutually cobounded subsets \mathcal{H}_b=\{H_{b,\alpha}\}_{\alpha\in \Lambda}.

3. For each b\in B the barycentre map \partial\widehat{F}_b\to\widehat{F}_b is uniformly coarsely surjective.

4. The fiber-identification maps are uniformly type preserving and satisfy qi-preserving electrocution condition.

5. The induced coned-off metric (graph) bundle satisfies a (\lambda_k, M_k,n_k)-flaring condition, for each k\geq 1.

6. Cone-bounded hallways strictly flare condition is satisfied.

Then X is hyperbolic relative to the family \mathcal{C} of maximal cone-subbundles of horosphere-like spaces. Suppose i:A\to B is a qi embedding and Y is the pullback of A under i. Then Y is hyperbolic relative to the family \mathcal{C}_Y of maximal cone-subbundles of horosphere-like spaces and (Y,\mathcal{C}_Y)\to (X,\mathcal{C}) admits the CT map.

[pranab-mahan, Lemma 1.1]

1. For every K_1,K_2\geq 1 and D\geq 0 there is K_{\ref{elem-lemma1}}=K_{\ref{elem-lemma1}}(K_1,K_2,D), such that the following holds: A K_1-coarsely Lipschitz map with a K_2-coarsely Lipschitz, D-coarse inverse is a K_{\ref{elem-lemma1}}-quasiisometry.

2. Given K\geq 1 and R\geq 0 there exists D_{\ref{elem-lemma1}}=D_{\ref{elem-lemma1}}(K,R) such that the following holds: Suppose X,Y are any two metric spaces and f:X\rightarrow Y is a K-quasiisometry which is R-coarsely surjective. Then there is a K_{\ref{elem-lemma1}}-quasiisometric D_{\ref{elem-lemma1}}-coarse inverse of f.

[pranab-mahan, Proposition 2.10] Suppose X is a geodesic metric space and Y is any metric space. A map f:X\to Y is coarsely C-Lipschitz for some C\geq 0 if for all x_1, x_2\in X, d_X(x_1,x_2)\leq 1 implies d_Y(f(x_1),f(x_2))\leq C.

Given K\geq 1,D\geq 0, we have K_{\ref{extend quasigeodesic}}=K_{\ref{extend quasigeodesic}}(K,D) such that following holds. Let (X,d) be a geodesic space and let \alpha be a K-quasigeodesic in X joining x_1,x_2\in X. Let y_1,y_2\in X such that d(x_i,y_i)\leq D, for i=1,2. Let \alpha' be the concatenation of \alpha and the geodesics [x_1,y_1] and [x_2,y_2]. Then \alpha' is a K_{\ref{extend quasigeodesic}}-quasigeodesic in X.

[krishna-sardar, Lemma 2.8] (1) Y is a (connected) metric graph.
(2) The maps \psi_X and \phi_X|_{V(Y)} are coarsely 1-surjective, (1,1)-quasiisometries.
(3) The map \phi_X is a (1,3)-quasiisometry and it is a 1-coarse inverse of \psi_X.

(Stability of quasigeodesics, [GhH]) For all \delta\geq 0 and k\geq 1, \epsilon\geq 0 there is a constant D_{\ref{stab-qg}}=D_{\ref{stab-qg}}(\delta, k, \epsilon) such that the following holds:
Suppose Y is a \delta-hyperbolic geodesic metric space. Then the Hausdorff distance between a geodesic and a (k,\epsilon)-quasigeodesic joining the same pair of end points is less than or equal to D_{\ref{stab-qg}}.

[krishna-sardar, Lemma 2.34] Given K\geq 0, R\geq 0, \delta\geq 0 there is a constant D=D_{\ref{trivial lemma}}(R,K,\delta) such that the following holds:
Suppose X is a \delta-hyperbolic metric space and A is a K-quasiconvex subset of X. Suppose x,y\in X and \bar{x}, \bar{y}\in A respectively are their 1-approximate nearest point projections on A. Suppose z\in [x,y] and \bar{z} is a 1-approximate nearest point projection of z on A satisfying d(z,\bar{z})\leq R. Then d(z, [\bar{x},\bar{y}])\leq D.

[mitra-trees, Lemma 3.1] Given \delta\geq 0,C\geq 1, there exist D_{\ref{big npp small nbd}}, M_{\ref{big npp small nbd}} such that the following holds: Let Y be C-quasiconvex subsets of a \delta-hyperbolic metric space (X,d). For any pair of points z,z', if d(P_Y(z),P_Y(z'))>D_{\ref{big npp small nbd}}, then [P_Y(z),P_Y(z')]\subset N_{M_{\ref{big npp small nbd}}}([z,z']).

[pal-singh],[ps-kap, Theorem 9.28] Let \hat{\alpha} be an electric quasigeodesic in \widehat{X}. Then the corresponding electro-ambient path \alpha is a K_{\ref{electro ambient is quasigeodesic}}-quasigeodesic in X^h.

Let K\geq 1. Let h\in H and \sigma_1 be a K-quasigeodesic (ray) in X^h with h as an endpoint. Let \sigma_2 be the vertical ray \{(h,t)\mid t\geq 0\} lying in H^h. Then the concatenation \sigma=\sigma_1\ast\sigma_2 is a (2K+1)-quasigeodesic ray in X^h.

The path \lambda(a,b) is a K_{\ref{quasigeodesics}}-quasigeodesic in H^h.

There exists f_{\ref{proper embedding of cones}}:[0,\infty)\to[0,\infty) satisfying the following. The inclusion H\to H^h is a proper embedding as measured by f_{\ref{proper embedding of cones}}.

There exists f_{\ref{proper embedding in cone space}}:[0,\infty)\to[0,\infty) satisfying the following. The inclusion X\to X^h is a proper embedding as measured by f_{\ref{proper embedding in cone space}}.

Given R>0, there exists K_{\ref{lemma from limint}}=K_{\ref{lemma from limint}}(R)>0 such that the following holds. Let X be a geodesic metric space strongly hyperbolic relative to a collection of uniformly mutually cobounded subsets {\mathcal H} = \{H_{\alpha}\}_{\alpha \in \Lambda}. Let \gamma be a geodesic in X^h with endpoints lying outside the horoball-like sets. If x\in X such that x\in N_R^{X^h}(\gamma), then x\in N^X_{K_{\ref{lemma from limint}}}(\gamma\cap X).

[farb-relhyp, Proposition 4.6] Given \delta,C,D>0, there exists \Delta such that if X is a \delta-hyperbolic metric space with a collection {\mathcal H} of C-quasiconvex sets, then \widehat{X} is \Delta-hyperbolic.

[farb-relhyp] (1) Electric geodesics in {\mathcal E}(X,{\mathcal H}) and relative geodesics in X joining the same pair of points in X have similar intersection patterns with H for all H\in{\mathcal H}, i.e. they track each other off horosphere-like sets.
(2) Electric geodesics in {\mathcal E}(X,{\mathcal H}) (after identifying with {\mathcal E}({\mathcal G}(X,{\mathcal H}),{\mathcal H}^h)), and hyperbolic geodesics in {\mathcal G}(X,{\mathcal H}) joining the same pair of points in {\mathcal G}(X,{\mathcal H}) have similar intersection patterns with H^h for all H^h\in{\mathcal H}^h, i.e. they track each other off horoball-like sets.

[brahma-ibdd, farb-relhyp] Suppose X is a \delta-hyperbolic metric space with a collection {\mathcal H} of C-quasiconvex, D-mutually cobounded subsets. Then (X,{\mathcal H}) is strongly relatively hyperbolic.

[mahan-pal, Lemma 1.16] Given \delta>0, there exists P_{\ref{electric proj coarsely unique}}>0 depending \delta, D_{\ref{big npp small nbd}} and M_{\ref{big npp small nbd}} such that the following holds.
For H\in{\mathcal H}, x,y\in H and a geodesic {\hat{\alpha}} in \widehat{X}, we have d_{\widehat{X}}(i({\pi}_{\alpha}(x)),i({\pi}_{\alpha}(y)) \leq P_{\ref{electric proj coarsely unique}}.

[mahan-pal, Lemma 1.17] Given \delta>0, there exists P_{\ref{elec proj bound}}=P_{\ref{elec proj bound}}(\delta) such that the following holds. For x,y\in X and an electric geodesic \hat{\lambda} in \widehat{X}, we have d_{\widehat{X}}(\hat{\pi}_{\hat{\alpha}}(x)),\hat{\pi}_{\hat{\alpha}}(y))\leq P_{\ref{elec proj bound}}d_{\widehat{X}}(x,y)+P_{\ref{elec proj bound}}.

[mahan-pal, Lemma 1.18] Given \delta>0, K\geq 1, there exists P_{\ref{almost commute}}=P_{\ref{almost commute}}(\delta,K) such that the following holds.
Suppose \phi:X\to X is a type-preserving K-quasiisometry. Let a,b\in\widehat{X} and \widehat{\mu}_1 be a quasigeodesic in \widehat{X} joining a,b. Suppose the induced map \hat{\phi}:\widehat{X}\to\widehat{X} is also a quasiisometry. Let \widehat{\mu}_2 be a quasigeodesic in \widehat{X} joining \hat{\phi}(a) and \hat{\phi}(b). If x\in\widehat{X}, then d_{\widehat{X}}(\hat{\pi}_{\hat{\mu}_2}(\hat{\phi}(x)),\hat{\phi}(\hat{\pi}_{\hat{\mu}_1}(x)))\leq P_{\ref{almost commute}}.

[brahma-amalgeo, Lemma 7.1] The space (X_{pel},d_{pel}) is a hyperbolic metric space and the sets {\mathcal P}_\alpha are uniformly quasiconvex.

[mahan-reeves, Lemma 2.11] Given K,\epsilon >0, there exists C>0 such that the following holds:
Let \gamma_{pel} and \gamma^h denote respectively a (K,\epsilon)-partially electrocuted quasigeodesic in (X_{pel},d_{pel}) and a hyperbolic (K,\epsilon)-quasigeodesic in (X^h,d_h) joining a,b. Then \gamma_{pel} and \gamma^h have similar intersection patterns, i.e., \gamma^h\cap X lies in a (hyperbolic) C-neighborhood of (any representative of) \gamma_{pel}. Further, outside of a C-neighborhood of the horoballs that \gamma^h meets, \gamma^h and \gamma_{pel} track each other.

[Pal-thesis, Lemma 1.2.31] Let (X_i,{\mathcal H}_i) be strongly relatively hyperbolic metric spaces, for i=1,2. Let \phi:X_1\to X_2 be a type preserving quasiisometry. Then the map \phi^h:X_1^h\to X_2^h induced by \phi is a quasiisometry.

Let (X,d_X) be a geodesic metric space strongly hyperbolic relative to a collection of subsets {\mathcal H}'. Let Y be a metric graph approximation of X. Then Y is strongly hyperbolic relative to {\mathcal H}, where {\mathcal H}=\{\psi_X(H')\mid H'\in{\mathcal H}'\}.

[hamenstadt, Proposition 2.10] The boundary \partial X^h is homeomorphic to \partial\widehat{X}\cup\bigcup_{H\in{\mathcal H}}\partial H^h.

\partial\widehat{X} and \partial\widehat{X}_{pel} are homeomorphic.

[krishna-sardar, Lemma 2.45] Let x\in X be any point. Suppose \{x_n\} is any sequence in X and \beta_{m,n} is a k-quasigeodesic joining x_m to x_n for all m,n\in{\mathbb N}. Suppose \alpha_n is a k-quasigeodesic joining x to x_n. Then
(1) \{x_n\}\in \mathcal S if and only if \lim_{m,n\rightarrow\infty} d(x,\beta_{m,n})=\infty if and only if there is a constant D such that for all M>0 there is N>0 with Hd(\alpha_m\cap B(x;M),\alpha_n\cap B(x;M))\leq D for all m,n\geq N.
(2) Suppose, moreover, \xi\in \partial_s X and \gamma_n is a k-quasigeodesic in X joining x_n to \xi for all n\in {\mathbb N} and \alpha is a k-quasigeodesic joining x to \xi. Then x_n\to\xi if and only if d(x,\gamma_n)\to\infty if and only if there is constant D>0 such that for all M>0 there is N>0 with Hd(\alpha\cap B(x;M),\alpha_n\cap B(x;M))\leq D for all n\geq N.

[krishna-sardar, Lemma 2.50] Let x\in X be any point. Suppose \{\xi_n\} is any sequence of points in \partial_s X. Suppose \beta_{m,n} is a k-quasigeodesic line joining \xi_m to \xi_n for all m, n\in {\mathbb N} and \alpha_n is a k-quasigeodesic ray joining x to \xi_n for all n\in {\mathbb N}. Then
(1) \lim_{m,n\to\infty} d(x,\beta_{m,n})= \infty if and only if there is a constant D=D(k,\delta) such that for all M>0 there is N>0 with Hd(\alpha_m\cap B(x;M),\alpha_n\cap B(x;M))\leq D for all m,n\geq N and in this case, \{\xi_n\} converges to some point of \partial_s X.
(2) Suppose, moreover, \xi\in\partial_s X, \gamma_n is a k-quasigeodesic ray in X joining \xi_n to \xi for all n, and \alpha is a k-quasigeodesic ray joining x to \xi. Then \xi_n\rightarrow \xi if and only if d(x,\gamma_n)\to\infty if and only if there is constant D'=D'(k,\delta) such that for all M>0 there is N>0 with Hd(\alpha\cap B(x;M),\alpha_n\cap B(x;M))\leq D for all n\geq N. In this case \lim_{m,n\to\infty} d(x,\beta_{m,n})= \infty.

[krishna-sardar, Lemma 2.60] Suppose X,Y are hyperbolic metric spaces and f:Y\to X is a metrically proper map that admits CT. Then we have \Lambda(f(Y))=\partial f(\partial Y) if either Y is a proper metric space or f is a qi embedding.

(Ideal triangles are slim)[krishna-sardar, Proposition 2.43] Suppose X is a \delta-hyperbolic metric space. for x,y,z\in\overline{X}, a k-quasigeodesic triangle formed by joining each pair of points from \{x,y,z\} by k-quasigeodesics is an R=R_{\ref{ideal triangles are slim}}(\delta, k)-slim triangle.
In particular, if \gamma_1,\gamma_2 are two k-quasigeodesic rays with \gamma_1(0)=\gamma_2(0) and \gamma_1(\infty)=\gamma_2(\infty), then Hd(\gamma_1, \gamma_2)\leq R.

Let X be a geodesic metric space, x,y\in\partial X (resp. x,y\in X), let \gamma be a k-quasigeodesic in X joining x,y and \alpha:{\mathbb R}\to X (resp. \alpha:[0,l]\to X) be a (dotted) coarsely L-Lipschitz path joining x,y. If \alpha is a proper embedding as measured by a function f:[0,\infty)\to[0,\infty) and there exists D>0 such that Hd(\alpha,\gamma)\leq D, then \alpha is (dotted) K_{\ref{quasigeod criteria}}=K_{\ref{quasigeod criteria}}(k,f,D,L)-quasigeodesic in X.

(Barycenters of ideal triangles, [krishna-sardar, Lemma 2.46]) There exists r_0>0 such that given a hyperbolic geodesic space X, distinct points x,y,z\in\overline{X}, there exists an r_0-barycenter x_0\in X of a k-quasigeodesic triangle with vertices x,y,z. Moreover, there exists L_0\geq 0 such that any two barycenters are at most L_0 distance apart.

(Mitra’s criterion, [mitra-trees, Lemma 2.1]) Suppose X, Y are hyperbolic geodesic spaces and f:Y\to X is a proper embedding. Then f admits CT if the following holds:
(*) Let y_0\in Y. There exists a function \tau:{\mathbb R}_{\geq 0}\rightarrow {\mathbb R}_{\geq 0}, with the property that \tau(n)\rightarrow \infty as n\rightarrow \infty such that for all geodesic segments [y_1,y_2]_Y lying outside the n-ball around y_0\in Y, any geodesic segment [f(y_1),f(y_2)]_X lies outside the \tau(n)-ball around f(y_0)\in X.

(Mitra’s criterion for relatively hyperbolic spaces) Let X,Y be geodesic spaces strongly hyperbolic relative to a collection of subsets {\mathcal H}_X,{\mathcal H}_Y respectively. Let f:(X,{\mathcal H}_X)\to(Y,{\mathcal H}_Y) be a type preserving proper embedding. Then f admits CT if the following holds:
(**) Let x_0\in X. There exists a function \tau:{\mathbb R}_{\geq 0}\rightarrow {\mathbb R}_{\geq 0}, with the property that \tau(n)\to\infty as n\to\infty such that for x,y\in X and any partially electrocuted geodesic \alpha_{pel} joining x,y in X_{pel}, if \alpha_{pel}\cap X lies outside the n-ball around x_0 in X, then for any partially electrocuted geodesic \beta_{pel} joining f(x),f(y) in Y_{pel}, \beta_{pel}\cap Y lies outside the \tau(n)-ball around f(x_0) in Y.

(Path lifting lemma-I)[krishna-sardar, Lemma 3.8] Let \pi:X\to B be an (\eta,c)-metric bundle or an \eta-metric graph bundle. Suppose b_1,b_2\in B. Suppose \gamma:[0,L]\to B is a continuous, rectifiable, arc length parametrized path (resp. an edge path) in B joining b_1 to b_2. For any x\in F_{b_1}, there is a path \tilde{\gamma} in \pi^{-1}(\gamma) such that l(\tilde{\gamma})\leq(L+1)c (resp l(\tilde{\gamma})=L) joining x to some point of F_{b_2}.
In particular,if \gamma is a geodesic in B, then \tilde{\gamma} is a geodesic. Further, if \gamma is a (k,\epsilon)-(dotted) quasigeodesic, for k\geq 1 and \epsilon\geq 0, then \tilde{\gamma} is a c.(k+\epsilon+1)-qi lift. (Here, c=1 in the metric graph bundle case).

[krishna-sardar, Corollary 3.9] Given any metric (graph) bundle \pi:X\to B and b_1, b_2\in B we can define a map \phi:F_{b_1}\to F_{b_2} such that d_X(x, \phi(x))\leq 3c +3cd_B(b_1, b_2) (resp. d(x, \phi(x))=d_B(b_1,b_2)) for all x\in F_{b_1}.

[krishna-sardar, Lemma 3.10] Suppose \pi:X\to B is an (\eta,c)-metric bundle or \eta-metric graph bundle. Let R\geq 0 and b_1,b_2\in B. Then we have the following.
(1) Hd(F_{b_1},F_{b_2})\leq 3c+ 3cd_B(b_1,b_2) (resp. Hd(F_{b_1},F_{b_2})= d_B(b_1,b_2)).
(2) Suppose \phi_{b_1b_2}:F_{b_1}\to F_{b_2} is a map such that d(x,\phi_{b_1b_2}(x))\leq R for all x\in F_{b_1}. Then \phi_{b_1b_2} is a D_{\ref{fibers qi}}-surjective, K_{\ref{fibers qi}}=K_{\ref{fibers qi}}(R)-quasiisometry.
(3) For a map \psi_{b_1b_2}:F_{b_1}\to F_{b_2} with d(x,\psi_{b_1b_2}(x))\leq R' for all x\in F_{b_1}, we have d(\phi_{b_1b_2},\psi_{b_1b_2})\leq\eta(R+R'). In particular, the maps \phi_{b_1b_2} are coarsely unique.

[krishna-sardar, Corollary 3.11] A fiber identification map \phi_{b_1b_2}:F_{b_1}\to F_{b_2} satisfying d_X(x,\phi_{b_1b_2}(x))\leq R for all x\in F_{b_1} is a K_{\ref{fibers unif qi}}=K_{\ref{fibers unif qi}}(R)-quasiisometry.

[pranab-mahan, Corollaries 1.14, 1.16] (Bounded flaring condition) For all k\in{\mathbb R}, k\geq 1 there exists a function \mu_k:{\mathbb N}\to{\mathbb N} such that the following holds:
Suppose \pi: X\to B is an (\eta,c)-metric bundle or an \eta-metric graph bundle. Let \gamma\subset B be a a geodesic joining b_1,b_2\in B, and let \tilde{\gamma_1}, \tilde{\gamma_2} be two k-qi lifts of \gamma in X. Suppose \tilde{\gamma}_i(b_1)=x_i\in F_{b_1} and \tilde{\gamma}_i(b_2)=y_i\in F_{b_2}, i=1,2. Then f d_B(b_1,b_2)\leq N, we have d_{b_2}(y_1,y_2)\leq \mu_k(N) \mbox{max}\{d_{b_1}(x_1,x_2),1\}.

(Restriction bundle) Suppose \pi:X\to B is a metric (graph) bundle and A\subset B is a connected subset (resp. A is a connected subgraph) such that any pair of points in A can be joined by a path of finite length in A. Then the restriction of \pi to Y=\pi^{-1}(A) gives a metric (graph) bundle with the same parameters as that of \pi: X\to B where A and Y are given the induced metrics from B and X respectively.
Moreover, if f:Y\to X and g:A\to B are the inclusion maps then (f,g):(Y,A)\to (X,B) is a metric (graph) bundle morphism.

[krishna-sardar, Lemma 3.20] Given L\geq 0 and \phi_1,\phi_2:[0,\infty)\to[0,\infty) there exists \phi:[0,\infty)\to[0,\infty) such that the following hold:
Suppose we have the following commutative diagram (Figure 2) of maps between metric spaces satisfying the properties (1)-(3) below.

(1) All the maps (except possibly f') are coarsely L-Lipschitz.
(2) If d_{B_1}(b,b')\leq N, then Hd(\pi^{-1}_1(b), \pi^{-1}_1(b'))\leq\phi_1(N) for b,b'\in B_1 and N\geq 0.
(3) The restriction of f on the fibers of \pi_1 are uniformly properly embedded as measured by \phi_2.
Then d_Y(y,y')\leq R implies d_{X_1}(f'(y),f'(y'))\leq\phi(R) for all y,y'\in Y and R\in[0,\infty). In particular, if Y is a geodesic space or the vertex set of a connected metric graph with restricted metric, then f' is coarsely \phi(1)-Lipschitz.
Moreover, f' is coarsely unique, i.e. there is a constant D>0 such that if another map f'':Y\to X_1 makes the above diagram commutative then d(f',f'')\leq D.

(Pullbacks of metric bundles)[krishna-sardar, Proposition 3.19, Proposition 3.21] Suppose (X,B,\pi) is a metric bundle and g:B_1\to B is a Lipschitz map. Then there is a pullback. Moreover, it is coarsely unique up to a metric bundle isomorphism.

(Pullbacks for metric graph bundles)[krishna-sardar, Proposition 3.23] Suppose (X,B,\pi) is an \eta-metric graph bundle, B_1 is a metric graph and g:V(B_1)\to V(B) is a coarsely L-Lipschitz map for some constant L\geq 1. Then there exists a pullback \pi_1:X_1\to B_1 of g such that all the fiber maps f_b:\pi^{-1}_1(b)\to \pi^{-1}(g(b)), b\in V(B_1) are isometries with respect to induced length metrics from X_1 and X respectively and it is coarsely unique up to a metric graph bundle isomorphism.

[krishna-sardar, Corollary 3.24] Suppose \pi:X\to B is a metric graph bundle. Suppose A is a connected subgraph of B. Let g:A\to B denote the inclusion map. Let X_A=\pi^{-1}(A), \pi_A be the restriction of \pi and let f:X_A\to X denote the inclusion map. Then X_A is the pullback of X under g.

Suppose \pi':X'\to B' is an (\eta,c)-metric bundle. Then there is a metric graph bundle \pi:X\to B along with quasiisometries \psi_B:B'\to B and \psi_X: X'\to X such that
(1) \pi\circ \psi_X=\psi_B\circ \pi', and
(2) for b\in B', \psi_X|_{\pi'^{-1}(b)}:\pi'^{-1}(b)\to\pi^{-1}(\psi_B(b)) is a uniform quasiisometry.
Moreover, the maps \psi_X, \psi_B have coarse inverses \phi_X, \phi_B respectively making the following diagram commutative:

Let (X',B',\pi'), (Y',A',p') be metric bundles and let (f,g) be a metric bundle morphism from Y' to X'. Let (X,B,\pi), (Y,A,p) be the metric graph bundles corresponding to (X',B',\pi'), (Y',A',p') respectively. Then
(1) Let \bar{f}:=\psi_X\circ f\circ\phi_Y and \bar{g}:=\psi_B\circ g\circ\phi_A. Then (\bar{f},\bar{g}) gives a morphism of metric graph bundles from Y to X.
Moreover, if Y' is the pullback of X' under g and f is the pullback map, then Y is the pullback of X under \bar{g} and \bar{f} is the pullback map.
(2) If X',Y' are hyperbolic, then f admits a CT map if and only if so does \bar{f}.

(Path lifting lemma-II) Suppose \pi:X\to B is an \eta-relatively hyperbolic metric (graph) bundle. Let b_1, b_2\in B (resp. b_1,b_2\in V(B)). For any x\in F_{b_1}, a geodesic \gamma can be lifted to a geodesic \tilde{\gamma} joining x to some y\in F_{b_2}. In particular, if x\in H_{b_1,\alpha}, for some H_{b_1,\alpha}\in{\mathcal H}_{b_1}, then there exists D_{\ref{lifting geodesics-2}} depending on d_B(b_1,b_2) such that for every b\in\gamma, \tilde{\gamma}(b)\in N_{D_{\ref{lifting geodesics-2}}}(H_{b,\alpha}) for some H_{b,\alpha}\in{\mathcal H}_b.

Given \eta':(0,\infty)\to(0,\infty), there exists \eta:{\mathbb N}\to{\mathbb N} such that the following holds. Suppose (X',B',\pi') is an (\eta',c)-relatively hyperbolic metric bundle. Then (X,B,\pi) is an \eta-relatively hyperbolic metric graph bundle.

The space \mathcal{X} is a \eta_{\ref{new mbdl}}-metric graph bundle over B.

Let \pi_i:X_i\to B_i, i=1,2 be relatively hyperbolic metric graph bundles with {\mathcal H}_i=\bigcup_{b\in V(B_i)}{\mathcal H}_b. Let (f,g) be an R-type preserving morphism from (X_1,{\mathcal H}_1) to (X_2,{\mathcal H}_2), for some R>0. The induced map (\hat{f}:\mathcal{X}_1\to\mathcal{X}_2,g) is a metric graph bundle morphism.

Let \pi_i:X_i\to B_i, i=1,2 be relatively hyperbolic metric graph bundles. Let (f,g) be a type preserving isomorphism from (X_1,{\mathcal H}_1) to (X_2,{\mathcal H}_2). Then the induced map (\hat{f}:\mathcal{X}_1\to\mathcal{X}_2,g) is a metric graph bundle isomorphism.

Let (X,B,\pi) be a relatively hyperbolic metric graph bundle. Let A be a hyperbolic metric space and g:A\to B be a coarsely L-Lipschitz map. Then there exists a pullback and the pullback is a relatively hyperbolic metric graph bundle. Moreover, the pullback map is type preserving.

Let (X,B,\pi) be a relatively hyperbolic metric graph bundle and g:A\to B be a coarsely Lipschitz map. Let (Y,A,p) be the pullback of X under g and f be the pullback map. Then \mathcal{Y}:={\mathcal E}(Y,{\mathcal H}_Y) is the pullback of \mathcal{X}:={\mathcal E}(X,{\mathcal H}) under g with the induced map \hat{f}:\mathcal{Y}\to\mathcal{X} as the pullback map.

[pranab-mahan, Proposition 2.10, Proposition 2.12] (Global qi sections for metric (graph) bundles) For all \delta',c\geq 0, N\geq 0 and \eta:[0,\infty) \rightarrow [0,\infty) there exists K_{\ref{existence-qi-section}}=K_{\ref{existence-qi-section}}(c,\eta,\delta',N) such that the following holds (c=1 for metric graph bundles).
Suppose \pi:X'\rightarrow B' is an (\eta,c)-metric bundle or an \eta-metric graph bundle with fibers such that for each b\in B' (b\in V(B')),
(i) F_b is \delta'-hyperbolic with respect to the induced path metric.
(ii) The barycenter map \partial^3 F_b\to F_b is coarsely N-surjective.
Then there is a K_{\ref{existence-qi-section}}-qi section over B' through each point of X'.

For all \delta',c\geq 0, N\geq 0 and \eta:[0,\infty) \rightarrow [0,\infty) there exists K_{\ref{existence-qi-section-2}}=K_{\ref{existence-qi-section-2}}(c,\eta,\delta',N) such that the following holds.
Suppose \pi:X\rightarrow B is a \eta-relatively hyperbolic metric graph bundle such that for each b\in V(B), the barycenter map \partial^3\widehat{F}_b\to\widehat{F}_b is coarsely N-surjective. Then there is a K_{\ref{existence-qi-section-2}}-qi section in \mathcal{X} over B, through each point of X, such that it lies in X.

Given K\geq 0, \delta\geq 0 there exists C=C_{\ref{ladders are qi embedded}}(K)\geq 0, R= R_{\ref{ladders are qi embedded}}(K)\geq 0 and K_{\ref{ladders are qi embedded}}(\delta, K)\geq 0 such that the following holds. Let \pi:X\to B be a \eta-metric graph bundle and let {\Sigma}_1, {\Sigma}_2 be two K-qi sections in X and {\mathbb L}={\mathbb L}({\Sigma}_1, {\Sigma}_2). Then the following hold.
(1) The map \Pi:\mathcal{X}\to{\mathbb L} is a coarse C-Lipschitz retraction.
(2) Given a k-qi section \gamma in \mathcal{X} over a geodesic in B, \Pi(\gamma) is a (C+2kC)-qi section in \mathcal{X} contained in {\mathbb L} over the same geodesic in B.
(3) If the barycenter map of the fibers of \mathcal{X} are uniformly coarsely surjective, then through any point of {\mathbb L} there is (1+2K)C-qi section contained in {\mathbb L}.
(4) The R-neighbourhood of {\mathbb L} is connected and uniformly qi embedded in \mathcal{X}. In particular, if \mathcal{X} is \delta-hyperbolic then {\mathbb L} is K_{\ref{ladders are qi embedded}}(\delta, K)-quasiconvex in \mathcal{X}.

([pranab-mahan, Theorem 4.3 and Proposition 5.8]) Suppose \pi:X\to B is a metric (graph) bundle such that

1. B is a hyperbolic metric space.

2. Each fiber F_b, b\in B (resp. b\in V(B)) is a uniformly hyperbolic metric space.

3. For each fiber, the barycentre map \partial^3 F_b\to F_b is uniformly coarsely surjective.

Then X is a hyperbolic metric space if and only if a flaring condition is satisfied.

[pranab-mahan, Proposition 5.17] Let K=\pi_1(S^h) be the fundamental group of a surface with finitely many punctures and let K_1,\ldots,K_n be its peripheral subgroups. Let 1\to K\to G\stackrel{\pi}{\to} Q\to1 and 1\to K_i\to N_G(K_i)\stackrel{\pi}{\to}Q_i\to 1 be the induced short exact sequences of groups. Then G is strongly hyperbolic relative to \{N_G(K_i)\}^n_{i=1} if and only if Q is a convex cocompact subgroup of the pure mapping class group of S^h.

Let \pi:X\rightarrow B be an \eta-metric (graph) bundle such that

1. (C1) B is a \delta_0-hyperbolic metric space.

2. (C2) For each b\in V(B), the fiber F_b strongly hyperbolic relative to a collection of uniformly mutually cobounded subsets {\mathcal H}_b.

3. (C3) For each b\in V(B), the barycentre map \partial^3\widehat{F}_b\to \widehat{F}_b is coarsely N-surjective.

4. (C4) The induced coned-off metric (graph) bundle satisfies the (\nu_k, M_k,n_k)-flaring condition, for each k\geq 1.

Then X is weakly hyperbolic relative to {\mathcal C}.

Suppose the metric (graph) bundle \mathcal{X} is hyperbolic and it satisfies a cone-bounded strictly flare condition. Then there exists D>0 such that {\mathcal B} is a mutually D-cobounded collection of subsets.

Let \pi:X\rightarrow B be an \eta-metric (graph) bundle such that

1. (C1) B is a \delta_0-hyperbolic metric space.

2. (C2) Each fiber F_b is strongly hyperbolic relative to a collection of uniformly mutually cobounded subsets {\mathcal H}_b.

3. (C3) For each fiber, the barycentre map \partial^3\widehat{F}_b\to \widehat{F}_b is coarsely N-surjective.

4. (C4) The induced coned-off metric graph bundle satisfies the (\nu_k, M_k,n_k)-flaring condition, for each k\geq 1.

5. (C5) The cone bounded strictly flare condition is satisfied.

Then X is strongly hyperbolic relative to {\mathcal C}.

Let \pi:X\rightarrow B be an \eta-metric (graph) bundle such that

1. (C1) B is a \delta_0-hyperbolic metric space.

2. (C2) For each b\in V(B), the fiber F_b strongly hyperbolic relative to a collection of uniformly separated closed subsets {\mathcal H}_b.

3. (C3) For each b\in V(B), the barycentre map \partial^3\widehat{F}_b\to \widehat{F}_b is coarsely N-surjective.

If (X,{\mathcal C}) is strongly relatively hyperbolic, then the induced coned-off metric (graph) bundle satisfies a flaring condition and a cone bounded strictly flare condition.

[krishna-sardar, Theorem 5.2] Suppose X is a metric (graph) bundle over a hyperbolic metric space B such that X is hyperbolic and all the fibers are nonelementary uniformly hyperbolic, and the barycenter map is uniformly coarsely surjective. Suppose i:A\rightarrow B is a Lipschitz, qi embedding and Y is the pullback of X under i. Then Y is hyperbolic and f:Y\to X admits the CT map.

The metric graph bundle (X,B,\pi) is strongly hyperbolic relative to {\mathcal C}=\{\psi_X({\mathcal C}'_{\alpha})\mid{\mathcal C}'_{\alpha}\in{\mathcal C}'\}.

The map \psi^h:X'^h\to X^h induced by \psi is a quasiisometry.

The metric graph bundle (Y,A,p) is strongly hyperbolic relative to a collection of subsets {\mathcal C}_Y.

Suppose \pi:X\rightarrow B is a metric (graph) bundle satisfying the hypotheses of Theorem 1.2. Let i: A\rightarrow B is a Lipschitz k-qi embedding and p:Y\rightarrow A be the pullback bundle with the pullback map f:Y\rightarrow X. Then \mathcal{Y} is a pullback of \mathcal{X}.

[krishna-sardar, Lemma 5.20] The pullback of a metric (graph) bundle is metrically properly embedded in it.

[krishna-sardar, Corollary 5.10] There exists constants R_0>0, R_1, R'_1, D_1 and there is a partition 0=t_0<t_1<\cdots <t_n=l of [0,l] and K_1-qi sections {\Sigma}_i passing through \alpha(t_i), 0\leq i\leq n inside {\mathbb L}({\Sigma}, {\Sigma}') such that the following holds.
(1) {\Sigma}_0={\Sigma}, {\Sigma}_n={\Sigma}'.
(2) For 0\leq i\leq n-2, {\Sigma}_{i+1}\subset {\mathbb L}({\Sigma}_i, {\Sigma}').
(3) For 0\leq i\leq n-2, there are two possibilities. (\textup{I}) d_h({\Sigma}_i, {\Sigma}_{i+1})=R_0,
(\textup{II}) d_h({\Sigma}_i, {\Sigma}_{i+1})>R_0 and there is a K_2-qi section {\Sigma}'_i through \alpha(t_{i+1}-1) inside {\mathbb L}({\Sigma}_i, {\Sigma}_{i+1}) such that d_h({\Sigma}_i, {\Sigma}'_i)<R_1. In either case d_X({\Sigma}_i, {\Sigma}_{i+1})>R'_1 and {\Sigma}_i, {\Sigma}_{i+1} are D_1-cobounded in \mathcal{X}.
(4) d_h({\Sigma}_{n-1}, {\Sigma}_n)\leq R_0.

[krishna-sardar, Lemmas 5.12,5.18, Proposition 2.38] There exists K_{\ref{main quasigeodesics}}\geq1 such that the path c(y,y') (resp. \bar{c}(y,y')) is a K_{\ref{main quasigeodesics}}-quasigeodesic in \mathcal{X} (resp. \mathcal{Y}).

There exists K_{\ref{vertical qg}}\geq 1 such that for any x\in{\mathbb L}\cap X, there exists a vertical quasigeodesic ray r_x:V([\pi(x),b_0])\to X such that r_x(\pi(x))=x and d_B(\pi(x),v)\leq d_X(r_x(\pi(x)),r_x(v))\leq K_{\ref{vertical qg}}d_B(\pi(x),v).

[krishna-sardar, Lemma 5.21] Given D>0, there is D_{\ref{criterion satisfied}}>0 such that the following holds. If d_{\mathcal{X}}(y_0,c)\leq D then d_{\mathcal{Y}}(y_0,\bar{c})\leq D_{\ref{criterion satisfied}}.

Given D>0, there is D_{\ref{criterion satisfied-2}}=D_{\ref{criterion satisfied-2}}(D)>0 such that the following holds. If d_{X}(y_0,c\cap X)\leq D then d_{Y}(y_0,\bar{c}\cap Y)\leq D_{\ref{criterion satisfied-2}}.

[Pal, Theorem 3.11] Suppose we have a short exact sequence of pairs of finitely generated groups

such that K is strongly hyperbolic relative to a non-trivial proper subgroup K_1 and G preserves cusps. Moreover, if G is weakly hyperbolic relative to K_1 and strongly hyperbolic relative to N_G(K_1), then the embedding i:\Gamma_K\to\Gamma_G admits a CT map.

Let 1\to (K,K_1)\to (G,N_G(K_1))\to (Q,Q_1)\to 1 be a short exact sequence of pairs of finitely generated groups such that K is strongly hyperbolic relative to a proper non-trivial subgroup K_1 and suppose G preserves cusps. Let G be weakly hyperbolic relative to K_1 and a cone-bounded strictly flare condition be satisfied by {\mathcal E}(G,K_1). Then G is strongly relatively hyperbolic relative to N_G(K_1).

Suppose 1\to (K,K_1)\to (G,N_G(K_1))\to (Q,Q_1)\to 1 is a short exact sequence of relatively hyperbolic groups where K is strongly hyperbolic relative to a proper non-trivial subgroup K_1 and G preserves cusps. Suppose G is weakly hyperbolic relative to K_1 and it is strongly hyperbolic relative to N_G(K_1). Suppose Q' is a qi embedded subgroup of Q and G_1=\pi^{-1}(Q'). Then (G_1,N_{G_1}(K_1)) is strongly relatively hyperbolic and (G_1,N_{G_1}(K_1))\to(G,N_{G}(K_1)) admits CT.

\displaystyle{\partial \mathcal{X}=\Lambda\widehat{F}\cup\bigcup_{\xi\in\partial B}\partial^{\xi}\mathcal{X}}.

Suppose b\in B and \alpha_n is a sequence of geodesics in B joining b to b_n. Let \tilde{\alpha}_n is a uniform qi lift of \alpha_n for all n. Suppose \tilde{\alpha}_n joins \tilde{\alpha}_n(b)=x^0_n to \tilde{\alpha}_n(b_n)=x_n and suppose the set \{x^0_n\} has finite diameter. If x_n\to z\in \partial\mathcal{X}, then \lim_{n\to\infty}b_n exists. If \xi=\lim_{n\to\infty}b_n and \alpha:[0,\infty)\to B is a geodesic ray joining b to \xi, then there is a uniform qi lift \tilde{\alpha} of \alpha such that \tilde{\alpha}(\infty)=z.

Let z\in\bigsqcup_{\xi\in\partial B}\partial^{\xi}\mathcal{X}. Then there exists a qi lift \tilde{\alpha} in \mathcal{X} such that \tilde{\alpha}\subset X and \tilde{\alpha}(\infty)=z in \partial\mathcal{X}.

[krishna-sardar, Lemma 6.14] Suppose {\mathbb L} is an infinite girth ladder.

1. There is a uniformly coarsely Lipschitz retraction \Pi:\mathcal{X}\rightarrow {\mathbb L} such that for all b\in B and x\in \widehat{F}_b, \Pi(x) is an electric projection of x in \widehat{F}_b on {\mathbb L}\cap \widehat{F}_b.

2. Infinite girth ladders are uniformly quasiconvex and there exists C>0 such that a C-neighbourhood of the ladder is qi embedded in \mathcal{X}.

3. Through any point of {\mathbb L}, there exists a \bar{K}_0-qi section contained in {\mathbb L}.

4. Any qi section in {\mathbb L} coarsely bisects it into two subladders.

[krishna-sardar, Lemma 6.19]. Let {\mathbb L}={\mathbb L}(z_1;z_2) for (z_1,z_2)\in\partial^{(2)}_{\mathcal{X}}(\widehat{F}). Let \xi\in\partial B and \beta:[0,\infty)\to B be a continuous, arc length parameterized \kappa_0-quasigeodesic in B. Then any two qi lifts of \beta in {\mathbb L} are asymptotic in \mathcal{X}.

Suppose we have a metric graph bundle satisfying the hypotheses of Theorem 1.2. Suppose \gamma is a quasigeodesic line in \mathcal{Y} such that \gamma(\infty),\gamma(-\infty) are identified in \partial\mathcal{X}. Then \gamma(\pm\infty)\in\Lambda_{\mathcal{Y}}(\widehat{F}_b)\setminus\bigcup_{\xi\in\partial A}\partial^{\xi}\mathcal{Y} for some (any) b\in V(A).

Suppose we have a metric graph bundle satisfying the hypotheses of Theorem 1.2 such that X is a proper metric space. Let \gamma be a geodesic line in Y^h with \partial i_{Y,X}(\gamma(\infty))=\partial i_{Y,X}(\gamma(-\infty)) in \partial X^h. Let b\in V(A) and F=F_b. Then
(1) \gamma(\pm\infty)\in\partial i_{F,Y}(\partial F^h).

(2) There exists a uniform quasigeodesic line \sigma in F^h at a finite Hausdorff distance from \gamma, in Y^h, satisfying \partial i_{F,Y}(\sigma(\pm\infty))=\gamma(\pm\infty).

\ensuremath{\partial^{(2)}_{X}}(F)=\displaystyle{\big(\cup_{\xi\in\partial B}\ensuremath{\partial^{(2)}_{\xi}}(F)\big)^*}.

Suppose we have the hypothesis of Theorem 1.2 such that the fibers of the bundle are proper metric spaces. Let b\in A. Suppose the CT map \partial i:\partial F_b^h\to\partial X^h is surjective. Then the CT map \partial i_Y:\partial F_b^h\to\partial Y^h is also surjective.

Suppose 1\to (K,K_1)\to (G,N_G(K_1))\to (Q,Q_1)\to 1 is a short exact sequence as in Theorem 6.2. Suppose Q'< Q is qi embedded and G_1=\pi^{-1}(Q'). Then the CT map \partial K^h\to\partial G_1^h is surjective.

[mj-rafi, Theorem 1.1] Let 1\to K\to G\stackrel{\pi}{\to} Q\to 1 be a short exact sequence of hyperbolic groups, where K is a (closed) surface group or a free group and Q is convex cocompact in Teichmuller space or outer space respectively. Let H<K be a finitely generated infinite index subgroup of K. Then H is a quasiconvex subgroup in G.

[mj-rafi, Theorem 1.3] Let K=\pi_1(S^h) be the fundamental group of a surface with finitely many punctures and let K_1,\ldots,K_n be its peripheral subgroups. Let 1\to K\to G\stackrel{\pi}{\to} Q\to1 and 1\to K_i\to N_G(K_i)\stackrel{\pi}{\to}Q\to 1 be the induced short exact sequences of groups. Suppose Q is a convex cocompact subgroup of the pure mapping class group of S^h. Let K'<K be a finitely generated infinite index subgroup of K. Then K' is a relatively quasiconvex subgroup of G.

[sardar-mj, Theorem 4.3] Let Q_1<Q be a quasiconvex subgroup of Q and G_1=\pi^{-1}(Q_1). Let H_1<G_1 be an infinite index quasiconvex subgroup of G_1. Then H_1 is quasiconvex in G.

[sardar-mj, Proposition 2.8] Let G be a finitely presented group, H,K<G, where H is finitely generated. Let \Gamma=\Gamma(G,S) and \Gamma_1=\Gamma(H,S_1) be the Cayley graphs, where S,S_1 are finite generating sets of G,H respectively. Let D>0 and A\subset N_D(K)\cap H be an infinite subset. Suppose there exists r\geq 1 such that any pair of points in A can be joined by a path which lies in an r-neighbourhood of A in \Gamma_1. Then there exists a finitely generated infinite subgroup K'\subset H\cap K such that A/K' is finite. In particular, A is contained in a finite neighbourhood of K' in \Gamma.

[mj-rafi, Lemma 3.5] Let (G,{\mathcal H}) be a relatively hyperbolic group and (H,{\mathcal H}') be a relatively hyperbolic subgroup such that (H,{\mathcal H}')\to(G,{\mathcal H}) admits CT. Then H is relatively quasiconvex in G if and only if \partial^{(2)}_G(H)=\emptyset.

[mitraonss] Let (G,{\mathcal H}) be a relatively hyperbolic group and (H_1,{\mathcal H}_1) and (H_2,{\mathcal H}_2) be relatively hyperbolic subgroups such that H_2<H_1 and (H_2,{\mathcal H}_2)\stackrel{i}{\to}(H_1,{\mathcal H}_1) and (H_1,{\mathcal H}_1)\stackrel{j}{\to}(G,{\mathcal H}) admit CT map. Then \partial^{(2)}_G(H_2)=\partial^{(2)}_{H_1}(H_2)\cup\partial i^{-1}(\partial^{(2)}_{G}(H_1)).

Let {\mathcal K}=\{K_i\}_{i=1}^n,{\mathcal H}=\{N_G(K_i)\}_{i=1}^n. Let

be a short exact sequence of relatively hyperbolic groups. Let Q_1<Q be a quasiconvex subgroup of Q and let \pi^{-1}(Q_1)=G_1. Suppose H is an infinite index relatively quasiconvex subgroup of G_1. If H is not relatively quasiconvex in G, then there exists an infinite index relatively quasiconvex subgroup of K which is not relatively quasiconvex in G.

Let Q_1<Q be a quasiconvex subgroup of Q and let \pi^{-1}(Q_1)=G_1. Then any infinite index relatively quasiconvex subgroup of G_1 is a relatively quasiconvex subgroup of G.