We obtain explicit expressions for the annealed complexities associated respectively with the total number of (i) stationary points and (ii) local minima of the energy landscape for an elastic manifold with internal dimension d < 4 embedded in a random medium of dimension N ≫ 1 and confined by a parabolic potential with the curvature parameter µ. These complexities are found to both vanish at the critical value µc identified as the Larkin mass. For µ < µc the system is in complex phase corresponding to the replica symmetry breaking in its T = 0 thermodynamics. The complexities vanish respectively quadratically (stationary points) and cubically (minima) at µ − c . For d ≥ 1 they admit a finite "massless" limit µ = 0 which is used to provide an upper bound for the depinning threshold under an applied force.
PACS numbers:Numerous physical systems can be modeled by a collection of points or particles coupled by an elastic energy, usually called an elastic manifold, submitted to a random potential (see [1] for a review). They are often called "disordered elastic systems" and generically exhibit pinning in their statics and depinning transitions and avalanches in their driven dynamics [2][3][4][5][6].The manifold can be parameterized by a N -componentis the sum of an elastic energy, given by the (discrete) Laplacian matrix −t 0 ∆ xy , t 0 > 0, a quadratic confining energy controlled by the curvature parameter µ 0 > 0 (or, alternatively, the "mass" m = √ µ 0 ) and a centered Gaussian random potential with covarianceparametrized by a function B(z). This random potential is thus uncorrelated in the internal space and statistically translational invariant in the embedding space. We use periodic boundary conditions, i.e. the Laplacian eigenmodes are plane waves ∼ e ikx with eigenvalues ∆(k).Examples are ∆(k) = 2(cos k − 1) with k = 2πn/L, n = 0, ..L − 1 in d = 1, and for the continuum model ∆(k) = −k 2 in any dimension with k ∈ R d .The energy landscape provided by the functional (1) is complex and necessarily high-dimensional, i.e. involves many interacting and competing degrees of freedom, leading to glassy behavior. This necessitates to use methods and ideas of statistical mechanics of disordered systems such that Replica Symmetry Breaking (RSB), Functional Renormalization Group, etc. for understanding their properties [8][9][10][11]. Among other approaches, the problem of characterizing complexity of random high-dimensional landscapes by understanding statistical structure of their stationary points (minima, maxima and saddles), defined in our model by the condition δH δu(x) = 0 for all x, has attracted considerable intrinsic interest in recent years in pure and applied mathematics, see e.g. [12][13][14][15][16], as well as in theoretical physics, see [17][18][19][20][21][22][23][24] and references to earlier works in [25].For the energy functional (1)-(2) in the simplest 'toy model' limit d = 0 with no elastic interactions (when x is essentially a single point, L d = 1), the mean number of stationary points, N tot , an...