Mathematical aspects of Feynman integrals

In the present dissertation we consider Feynman integrals in the framework of dimensional regularization. As all such integrals can be expressed in terms of scalar integrals, we focus on this latter kind of integrals in their Feynman parametric representation and study their mathematical properties, partially applying graph theory, algebraic geometry and number theory. The three main topics are the graph theoretic properties of the Symanzik polynomials, the termination of the sector decomposition algorithm of Binoth and Heinrich and the arithmetic nature of the Laurent coefficients of Feynman integrals.rnrnThe integrand of an arbitrary dimensionally regularised, scalar Feynman integral can be expressed in terms of the two well-known Symanzik polynomials. We give a detailed review on the graph theoretic properties of these polynomials. Due to the matrix-tree-theorem the first of these polynomials can be constructed from the determinant of a minor of the generic Laplacian matrix of a graph. By use of a generalization of this theorem, the all-minors-matrix-tree theorem, we derive a new relation which furthermore relates the second Symanzik polynomial to the Laplacian matrix of a graph.rnrnStarting from the Feynman parametric parameterization, the sector decomposition algorithm of Binoth and Heinrich serves for the numerical evaluation of the Laurent coefficients of an arbitrary Feynman integral in the Euclidean momentum region. This widely used algorithm contains an iterated step, consisting of an appropriate decomposition of the domain of integration and the deformation of the resulting pieces. This procedure leads to a disentanglement of the overlapping singularities of the integral. By giving a counter-example we exhibit the problem, that this iterative step of the algorithm does not terminate for every possible case. We solve this problem by presenting an appropriate extension of the algorithm, which is guaranteed to terminate. This is achieved by mapping the iterative step to an abstract combinatorial problem, known as Hironaka's polyhedra game. We present a publicly available implementation of the improved algorithm. Furthermore we explain the relationship of the sector decomposition method with the resolution of singularities of a variety, given by a sequence of blow-ups, in algebraic geometry.rnrnMotivated by the connection between Feynman integrals and topics of algebraic geometry we consider the set of periods as defined by Kontsevich and Zagier. This special set of numbers contains the set of multiple zeta values and certain values of polylogarithms, which in turn are known to be present in results for Laurent coefficients of certain dimensionally regularized Feynman integrals. By use of the extended sector decomposition algorithm we prove a theorem which implies, that the Laurent coefficients of an arbitrary Feynman integral are periods if the masses and kinematical invariants take values in the Euclidean momentum region. The statement is formulated for an even more general class of integrals, allowing for an arbitrary number of polynomials in the integrand.

Knots and links in lens spaces

The aim of this dissertation is to improve the knowledge of knots and links in lens spaces. If the lens space L(p,q) is defined as a 3-ball with suitable boundary identifications, then a link in L(p,q) can be represented by a disk diagram, i.e. a regular projection of the link on a disk. In this contest, we obtain a complete finite set of Reidemeister-type moves establishing equivalence, up to ambient isotopy. Moreover, the connections of this new diagram with both grid and band diagrams for links in lens spaces are shown. A Wirtinger-type presentation for the group of the link and a diagrammatic method giving the first homology group are described. A class of twisted Alexander polynomials for links in lens spaces is computed, showing its correlation with Reidemeister torsion. One of the most important geometric invariants of links in lens spaces is the lift in 3-sphere of a link L in L(p,q), that is the counterimage of L under the universal covering of L(p,q). Starting from the disk diagram of the link, we obtain a diagram of the lift in the 3-sphere. Using this construction it is possible to find different knots and links in L(p,q) having equivalent lifts, hence we cannot distinguish different links in lens spaces only from their lift. The two final chapters investigate whether several existing invariants for links in lens spaces are essential, i.e. whether they may assume different values on links with equivalent lift. Namely, we consider the fundamental quandle, the group of the link, the twisted Alexander polynomials, the Kauffman Bracket Skein Module and an HOMFLY-PT-type invariant.

Monodromy calculations for some differential equations

In vielen Teilgebieten der Mathematik ist es w"{u}nschenswert, die Monodromiegruppe einer homogenen linearen Differenzialgleichung zu verstehen. Es sind nur wenige analytische Methoden zur Berechnung dieser Gruppe bekannt, daher entwickeln wir im ersten Teil dieser Arbeit eine numerische Methode zur Approximation ihrer Erzeuger.rnIm zweiten Abschnitt fassen wir die Grundlagen der Theorie der Uniformisierung Riemannscher Fl"achen und die der arithmetischen Fuchsschen Gruppen zusammen. Auss erdem erkl"aren wir, wie unsere numerische Methode bei der Bestimmung von uniformisierenden Differenzialgleichungen dienlich sein kann. F"ur arithmetische Fuchssche Gruppen mit zwei Erzeugern erhalten wir lokale Daten und freie Parameter von Lam'{e} Gleichungen, welche die zugeh"origen Riemannschen Fl"achen uniformisieren. rnIm dritten Teil geben wir einen kurzen Abriss zur homologischen Spiegelsymmetrie und f"uhren die $widehat{Gamma}$-Klasse ein. Wir erkl"aren wie diese genutzt werden kann, um eine Hodge-theoretische Version der Spiegelsymmetrie f"ur torische Varit"aten zu beweisen. Daraus gewinnen wir Vermutungen "uber die Monodromiegruppe $M$ von Picard-Fuchs Gleichungen von gewissen Familien $f:mathcal{X}rightarrow bbp^1$ von $n$-dimensionalen Calabi-Yau Variet"aten. Diese besagen erstens, dass bez"uglich einer nat"urlichen Basis die Monodromiematrizen in $M$ Eintr"age aus dem K"orper $bbq(zeta(2j+1)/(2 pi i)^{2j+1},j=1,ldots,lfloor (n-1)/2 rfloor)$ haben. Und zweitens, dass sich topologische Invarianten des Spiegelpartners einer generischen Faser von $f:mathcal{X}rightarrow bbp^1$ aus einem speziellen Element von $M$ rekonstruieren lassen. Schliess lich benutzen wir die im ersten Teil entwickelten Methoden zur Verifizierung dieser Vermutungen, vornehmlich in Hinblick auf Dimension drei. Dar"uber hinaus erstellen wir eine Liste von Kandidaten topologischer Invarianten von vermutlich existierenden dreidimensionalen Calabi-Yau Variet"aten mit $h^{1,1}=1$.

Heterotic T-folds with a small number of neutral moduli

We discuss non-geometric supersymmetric heterotic string models in D=4, in the framework of the free fermionic construction. We perform a systematic scan of models with four a priori left-right asymmetric Z2 projections and shifts. We analyze some 220 models, identifying 18 inequivalent classes and addressing variants generated by discrete torsions. They do not contain geometrical or trivial neutral moduli, apart from the dilaton. However, we show the existence of flat directions in the form of exactly marginal deformations and identify patterns of symmetry breaking where product gauge groups, realized at level one, are broken to their diagonal at higher level. We also describe an “inverse Gepner map” from Heterotic to Type II models that could be used, in certain non geometric settings, to define “effective” topological invariants.

Relating Crevassing to Non-Linear Strain in the Floating Part of Jakobshavn Isbrae, West Greenland

Jakobshavn Isbrae is a major ice stream that drains the west-central Greenland ice sheet and becomes afloat in Jakobshavn Isfiord (69degreesN, 49degreesW), where it has maintained the world's fastest-known sustained velocity and calving rate (7 km a(-1)) for at least four decades. The floating portion is approximately 12 km long and 6 km wide. Surface elevations and motion vectors were determined photogrammetrically for about 500 crevasses on the floating ice, and adjacent grounded ice, using aerial photographs obtained 2 weeks apart in July 1985. Surface strain rates were computed from a mesh of 399 quadrilateral elements having velocity measurements at each corner. It is shown that heavy crevassing of floating ice invalidates the assumptions of linear strain theory that (i) surface strain in the floating ice is homogeneous in both space and time, (ii) the squares and products of strain components are nil, and (iii) first- and second-order rotation components are small compared to strain components. Therefore, strain rates and rotation rates were also computed using non-linear strain theory. The percentage difference between computed linear and non-linear second invariants of strain rate per element were greatest (mostly in the range 40-70%) where crevassing is greatest. Isopleths of strain rate parallel and transverse to flow and elevation isopleths relate crevassing to known and inferred pinning points.

On the signature of positive braids and adjacency for torus knots

The objects of study in this thesis are knots. More precisely, positive braid knots, which include algebraic knots and torus knots. In the first part of this thesis, we compare two classical knot invariants - the genus g and the signature σ - for positive braid knots. Our main result on positive braid knots establishes a linear lower bound for the signature in terms of the genus. In the second part of the thesis, a positive braid approach is applied to the study of the local behavior of polynomial functions from the complex affine plane to the complex numbers. After endowing polynomial function germs with a suitable topology, the adjacency problem arises: for a fixed germ f, what classes of germs g can be found arbitrarily close to f? We introduce two purely topological notions of adjacency for knots and discuss connections to algebraic notions of adjacency and the adjacency problem.

Simulation of SLA-based VM-scaling algorithms for cloud-distributed applications

Cloud Computing has evolved to become an enabler for delivering access to large scale distributed applications running on managed network-connected computing systems. This makes possible hosting Distributed Enterprise Information Systems (dEISs) in cloud environments, while enforcing strict performance and quality of service requirements, defined using Service Level Agreements (SLAs). {SLAs} define the performance boundaries of distributed applications, and are enforced by a cloud management system (CMS) dynamically allocating the available computing resources to the cloud services. We present two novel VM-scaling algorithms focused on dEIS systems, which optimally detect most appropriate scaling conditions using performance-models of distributed applications derived from constant-workload benchmarks, together with SLA-specified performance constraints. We simulate the VM-scaling algorithms in a cloud simulator and compare against trace-based performance models of dEISs. We compare a total of three SLA-based VM-scaling algorithms (one using prediction mechanisms) based on a real-world application scenario involving a large variable number of users. Our results show that it is beneficial to use autoregressive predictive SLA-driven scaling algorithms in cloud management systems for guaranteeing performance invariants of distributed cloud applications, as opposed to using only reactive SLA-based VM-scaling algorithms.

Computational Schemes for Optimizing Domains of Attraction in Dynamical Systems

In this paper, several computational schemes are presented for the optimal tuning of the global behavior of nonlinear dynamical sys- tems. Specifically, the maximization of the size of domains of attraction associated with invariants in parametrized dynamical sys- tems is addressed. Cell Mapping (CM) tech- niques are used to estimate the size of the domains, and such size is then maximized via different optimization tools. First, a ge- netic algorithm is tested whose performance shows to be good for determining global maxima at the expense of high computa- tional cost. Secondly, an iterative scheme based on a Stochastic Approximation proce- dure (the Kiefer-Wolfowitz algorithm) is eval- uated showing acceptable performance at low cost. Finally, several schemes combining neu- ral network based estimations and optimiza- tion procedures are addressed with promising results. The performance of the methods is illus- trated with two applications: first on the well-known van der Pol equation with stan- dard parametrization, and second the tuning of a controller for saturated systems.

Abstract multiple specialization and its application to program parallelization.

Program specialization optimizes programs for known valúes of the input. It is often the case that the set of possible input valúes is unknown, or this set is infinite. However, a form of specialization can still be performed in such cases by means of abstract interpretation, specialization then being with respect to abstract valúes (substitutions), rather than concrete ones. We study the múltiple specialization of logic programs based on abstract interpretation. This involves in principie, and based on information from global analysis, generating several versions of a program predicate for different uses of such predicate, optimizing these versions, and, finally, producing a new, "multiply specialized" program. While múltiple specialization has received theoretical attention, little previous evidence exists on its practicality. In this paper we report on the incorporation of múltiple specialization in a parallelizing compiler and quantify its effects. A novel approach to the design and implementation of the specialization system is proposed. The resulting implementation techniques result in identical specializations to those of the best previously proposed techniques but require little or no modification of some existing abstract interpreters. Our results show that, using the proposed techniques, the resulting "abstract múltiple specialization" is indeed a relevant technique in practice. In particular, in the parallelizing compiler application, a good number of run-time tests are eliminated and invariants extracted automatically from loops, resulting generally in lower overheads and in several cases in increased speedups.

A technique for recursive invariance detection and selective program specialization

This paper presents a technique for achieving a class of optimizations related to the reduction of checks within cycles. The technique uses both Program Transformation and Abstract Interpretation. After a ñrst pass of an abstract interpreter which detects simple invariants, program transformation is used to build a hypothetical situation that simpliñes some predicates that should be executed within the cycle. This transformation implements the heuristic hypothesis that once conditional tests hold they may continué doing so recursively. Specialized versions of predicates are generated to detect and exploit those cases in which the invariance may hold. Abstract interpretation is then used again to verify the truth of such hypotheses and conñrm the proposed simpliñcation. This allows optimizations that go beyond those possible with only one pass of the abstract interpreter over the original program, as is normally the case. It also allows selective program specialization using a standard abstract interpreter not speciñcally designed for this purpose, thus simplifying the design of this already complex module of the compiler. In the paper, a class of programs amenable to such optimization is presented, along with some examples and an evaluation of the proposed techniques in some application áreas such as floundering detection and reducing run-time tests in automatic logic program parallelization. The analysis of the examples presented has been performed automatically by an implementation of the technique using existing abstract interpretation and program transformation tools.

Elastoplastic consolidation at finite strain. Part 2: finite element implementation and numerical examples

A mathematical model for finite strain elastoplastic consolidation of fully saturated soil media is implemented into a finite element program. The algorithmic treatment of finite strain elastoplasticity for the solid phase is based on multiplicative decomposition and is coupled with the algorithm for fluid flow via the Kirchhoff pore water pressure. A two-field mixed finite element formulation is employed in which the nodal solid displacements and the nodal pore water pressures are coupled via the linear momentum and mass balance equations. The constitutive model for the solid phase is represented by modified Cam—Clay theory formulated in the Kirchhoff principal stress space, and return mapping is carried out in the strain space defined by the invariants of the elastic logarithmic principal stretches. The constitutive model for fluid flow is represented by a generalized Darcy's law formulated with respect to the current configuration. The finite element model is fully amenable to exact linearization. Numerical examples with and without finite deformation effects are presented to demonstrate the impact of geometric nonlinearity on the predicted responses. The paper concludes with an assessment of the performance of the finite element consolidation model with respect to accuracy and numerical stability.

SMS freeforms for illumination

SMS 3D (simultaneous multiple surfaces in their three-dimensional version) is a well-known design method comprising two freeform surfaces that allow the perfect coupling of two wavefronts with another two. The design algorithm provides a collection of line pairs on both surfaces (called SMS spines), whose three-dimensional shape seems arbitrary at first sight. This paper shows that the shapes of the spines are partially governed by applying the étendue conservation theorem to the biparametric bundle of rays linking the paired spines, which is one lesser known étendue invariants found by Poincaré. The resulting formulae for the spines in three-dimensional space happen to coincide with the conventional étendue formulas of two-dimensional geometry, like for instance, the Hottel formula.

Métodos computacionales para visco-hiperelasticidad anisótropa en grandes deformaciones

El comportamiento mecánico de muchos materiales biológicos y poliméricos en grandes deformaciones se puede describir adecuadamente mediante formulaciones isocóricas hiperelásticas y viscoelásticas. Las ecuaciones de comportamiento elástico y viscoelástico y las formulaciones computacionales para materiales incompresibles isótropos en deformaciones finitas están ampliamente desarrolladas en la actualidad. Sin embargo, el desarrollo de modelos anisótropos no lineales y de sus correspondientes formulaciones computacionales sigue siendo un tema de investigación de gran interés. Cuando se consideran grandes deformaciones, existen muchas medidas de deformación disponibles con las que poder formular las ecuaciones de comportamiento. Los modelos en deformaciones cuadráticas facilitan la implementación en códigos de elementos finitos, ya que estas medidas surgen de forma natural en la formulación. No obstante, pueden dificultar la interpretación de los modelos y llevar a resultados pocos realistas. El uso de deformaciones logarítmicas permite el desarrollo de modelos más simples e intuitivos, aunque su formulación computacional debe ser adaptada a las exigencias del programa. Como punto de partida, en esta tesis se demuestra que las deformaciones logarítmicas representan la extensión natural de las deformaciones infinitesimales, tanto axiales como angulares, al campo de las grandes deformaciones. Este hecho permite explicar la simplicidad de las ecuaciones resultantes. Los modelos hiperelásticos predominantes en la actualidad están formulados en invariantes de deformaciones cuadráticas. Estos modelos, ya sean continuos o microestructurales, se caracterizan por tener una forma analítica predefinida. Su expresión definitiva se calcula mediante un ajuste de curvas a datos experimentales. Un modelo que no sigue esta metodología fue desarrollado por Sussman y Bathe. El modelo es sólo válido para isotropía y queda definido por una función de energía interpolada con splines, la cual reproduce los datos experimentales de forma exacta. En esta tesis se presenta su extensión a materiales transversalmente isótropos y ortótropos utilizando deformaciones logarítmicas. Asimismo, se define una nueva propiedad que las funciones de energía anisótropas deben satisfacer para que su convergencia al caso isótropo sea correcta. En visco-hiperelasticidad, aparte de las distintas funciones de energía disponibles, hay dos aproximaciones computational típicas basadas en variables internas. El modelo original de Simó está formulado en tensiones y es válido para materiales anisótropos, aunque sólo es adecuado para pequeñas desviaciones con respecto al equilibrio termodinámico. En cambio, el modelo basado en deformaciones de Reese y Govindjee permite grandes deformaciones no equilibradas pero es, en esencia, isótropo. Las formulaciones anisótropas en este último contexto son microestructurales y emplean el modelo isótropo para cada uno de los constituyentes. En esta tesis se presentan dos formulaciones fenomenológicas viscoelásticas definidas mediante funciones hiperelásticas anisótropas y válidas para grandes desviaciones con respecto al equilibrio termodinámico. El primero de los modelos está basado en la descomposición multiplicativa de Sidoroff y requiere un comportamiento viscoso isótropo. La formulación converge al modelo de Reese y Govindjee en el caso especial de isotropía elástica. El segundo modelo se define a partir de una descomposición multiplicativa inversa. Esta formulación está basada en una descripción co-rotacional del problema, es sustancialmente más compleja y puede dar lugar a tensores constitutivos ligeramente no simétricos. Sin embargo, su rango de aplicación es mucho mayor ya que permite un comportamiento anisótropo tanto elástico como viscoso. Varias simulaciones de elementos finitos muestran la gran versatilidad de estos modelos cuando se combinan con funciones hiperelásticas formadas por splines. ABSTRACT The mechanical behavior of many polymeric and biological materials may be properly modelled be means of isochoric hyperelastic and viscoelastic formulations. These materials may sustain large strains. The viscoelastic computational formulations for isotropic incompressible materials at large strains may be considered well established; for example Ogden’s hyperelastic function and the visco-hyperelastic model of Reese and Govindjee are well known models for isotropy. However, anisotropic models and computational procedures both for hyperelasticity and viscohyperelasticity are still under substantial research. Anisotropic hyperelastic models are typically based on structural invariants obtained from quadratic strain measures. These models may be microstructurallybased or phenomenological continuum formulations, and are characterized by a predefined analytical shape of the stored energy. The actual final expression of the stored energy depends on some material parameters which are obtained from an optimization algorithm, typically the Levenberg-Marquardt algorithm. We present in this work anisotropic spline-based hyperelastic stored energies in which the shape of the stored energy is obtained as part of the procedure and which (exactly in practice) replicates the experimental data. These stored energies are based on invariants obtained from logarithmic strain measures. These strain measures preserve the metric and the physical meaning of the trace and deviator operators and, hence, are interesting and meaningful for anisotropic formulations. Furthermore, the proposed stored energies may be formulated in order to have material-symmetries congruency both from a theoretical and from a numerical point of view, which are new properties that we define in this work. On the other hand, visco-hyperelastic formulations for anisotropic materials are typically based on internal stress-like variables following a procedure used by Sim´o. However, it can be shown that this procedure is not adequate for large deviations from thermodynamic equilibrium. In contrast, a formulation given by Reese and Govindjee is valid for arbitrarily large deviations from thermodynamic equilibrium but not for anisotropic stored energy functions. In this work we present two formulations for visco-hyperelasticity valid for anisotropic stored energies and large deviations from thermodynamic equilibrium. One of the formulations is based on the Sidoroff multiplicative decomposition and converges to the Reese and Govindjee formulation for the case of isotropy. However, the formulation is restricted to isotropy for the viscous component. The second formulation is based on a reversed multiplicative decomposition. This last formulation is substantially more complex and based on a corotational description of the problem. It can also result in a slightly nonsymmetric tangent. However, the formulation allows for anisotropy not only in the equilibrated and non-equilibrated stored energies, but also in the viscous behavior. Some examples show finite element implementation, versatility and interesting characteristics of the models.

Numerical issues in Lagrangian tracking and topological evolution of fluid particles in wallbounded turbulent flows

The determination of the local Lagrangian evolution of the flow topology in wall-bounded turbulence, and of the Lagrangian evolution associated with entrainment across the turbulent / non-turbulent interface into a turbulent boundary layer, require accurate tracking of a fluid particle and its local velocity gradients. This paper addresses the implementation of fluid-particle tracking in both a turbulent boundary layer direct numerical simulation and in a fully developed channel flow simulation. Determination of the sub-grid particle velocity is performed using both cubic B-spline, four-point Hermite spline and higher-order Hermite spline interpolation. Both wall-bounded flows show similar oscillations in the Lagrangian tracers of both velocity and velocity gradients, corresponding to the movement of particles across the boundaries of computational cells. While these oscillation in the particle velocity are relatively small and have negligible effect on the particle trajectories for time-steps of the order of CFL = 0.1, they appear to be the cause of significant oscillations in the evolution of the invariants of the velocity gradient tensor.

El Parque Agrario: estructura de preservación de los espacios agrarios en entornos urbanos en un contexto de cambio global

La propuesta del análisis de la figura de Parque Agrario en el ámbito español surge ante la constatación de que un nuevo paradigma está aconteciendo a escala estatal. Diversos focos se encuentran trabajando en paralelo, y de forma participada, en pos de la reformulación de las políticas públicas relacionadas con la agricultura periurbana. Estos focos ven en la figura de Parque Agrario un instrumento territorial que permite mejorar la sostenibilidad y cohesión territorial a través de la defensa de la gobernanza alimentaria local, sin olvidar la necesidad de conservación de los recursos naturales y el patrimonio paisajístico, junto a la prestación de múltiples servicios de los ecosistemas de estos ámbitos a la ciudadanía. Complementariamente, se empieza a vislumbrar el papel que esta figura puede desempeñar como herramienta de desarrollo territorial de los sistemas agrarios periurbanos, clave ante los efectos de carácter local que la globalización ejerce en estos territorios. La figura de Parque Agrario es una estructura que actúa bloqueando la base territorial, favoreciendo el desarrollo de la actividad agraria. Su mayor potencial es el de convertir el factor “proximidad urbana” de una amenaza a una oportunidad de desarrollo local endógeno que permita la continuidad de la agricultura, de los agricultores y del espacio agrario. La peculiaridad del Parque Agrario es que no es una figura al uso, estructurada y reglada por una legislación, sino que se trata de una iniciativa ad hoc, específica para cada caso, orientada a cumplir determinados objetivos de dinamización agraria, protección urbanística y valorización territorial. A pesar de la existencia de diversas definiciones y aportaciones sobre diferentes aspectos de la figura, no existe un análisis complejo de la misma en todas sus dimensiones, ni una tentativa de descripción de un modelo global y unitario del caso español y de sus potenciales resultados. Tampoco se han analizado en profundidad sus “invariantes” que se muestran como los elementos estructurantes del proyecto, capaces desarrollarse de forma diversa, de alcanzar diferentes niveles de complejidad, y de materializarse en función a las posibilidades que permita el marco normativo y legal. Por tanto, se plantea como objetivo principal de la tesis la definición de un modelo conceptual de Parque Agrario español, capaz de ser articulado e institucionalizado mediante un proceso de gobernanza, y que, como condición sine qua non sea duradero en el tiempo. Para poder llegar a describir un modelo colectivo se realiza, en primer lugar, un análisis genealógico que permita analizar sistemáticamente las propuestas desarrolladas en el ámbito español y los casos para establecer la existencia de una continuidad en la idea de Parque Agrario en las propuestas desarrolladas durante los últimos 25 años—sus invariantes—, y analice todos aquellos elementos que han ido enriqueciendo la figura en cada experiencia —sus variantes. Este análisis, además, ofrece como aportaciones el árbol genealógico y los mapas de dispersión de la figura y el primer catálogo de propuestas de Parque Agrario materializadas en proyecto. El resultado de la inducción de los datos obtenidos en el análisis genealógico es el modelo conceptual de Parque Agrario, que se define como una estructura orgánica de planificación-gestión-gobierno del territorio capaz de adaptarse a las necesidades específicas de todo sistema agrario periurbano que requiera la articulación-institucionalización de esta figura en él. Una vez descrito el modelo, se contrasta su fiabilidad mediante su aplicación como metodología de caracterización y evaluación de dos estudios de casos: uno exitoso, el Parque Agrario del Baix Llobregat, y uno frustrado, la propuesta de Parque Agrario de la Vega de Granada. ------------------------------------------------------ ABSTRACT -------------------------------------------------------------------- The proposed analysis of the figure of Agrarian Park in the Spanish sphere arises from the awareness that a new paradigm is happening at the state level. Different focuses are working in parallel, under participated programs, after the reformulation of public policies related to urban agriculture. These areas understand the figure of Agrarian Park as a territorial instrument for improving sustainability and territorial cohesion through the defense of local food governance, considering the need for conservation of natural resources and landscape heritage together with the multiple ecosystem services provided by these areas to the public. Additionally, the role that this figure can play as a tool for territorial development of peri-urban agrarian systems, which are key to the local effects that globalization has on these territories, is staring to be envisioned. The figure of Agrarian Park is a structure that works by blocking the territorial base to encourage the development of agrarian activity. Its greatest potential is to convert the threat of "urban proximity" into an opportunity for an endogenous local development that allows the persistence of agriculture, farmers and the agrarian space. The uniqueness of the Agrarian Park is that it is not a standard figure, structured and regulated by legislation, but rather an ad hoc initiative, specific to each case, designed to meet certain objectives of agrarian revitalization, urban protection and territorial enhance. Despite the existence of several definitions and contributions on different aspects of the figure, there is a lack of a complex analysis of it in all its dimensions, missing any attempt to describe a global and unitary model of the Spanish case and its potential outcomes. Its "invariants��� have neither been evaluated in depth, shown as the structural elements of the project able to be developed in different ways, to achieve numerous levels of complexity, and to be materialized according to the possibilities allowed by the regulatory and legal framework. Therefore, the definition of a conceptual model of Spanish Agrarian Park able to be articulated and institutionalized through a process of governance, and durable over time as a sine qua non requisite, it is proposed as the main aim of the thesis. To get to describe a collective model, a genealogical analysis that systematically analyzes the proposals and cases developed in the Spanish field is undertaken to verify the existence of a continuity of the idea of Agrarian Park on the proposals developed during the past 25 years -invariants-, and evaluate all the elements that have enriched this figure in each experience - variants. This analysis also provides as inputs a family tree, maps of dispersion of the figure and the first catalog of Agrarian Park proposals materialized into projects. The result of inducting the data obtained in the genealogical analysis is the Conceptual Model of Agrarian Park, defined as an organic planning-management-government structure of the territory able to adapt to the specific needs of all peri-urban agrarian systems that require the articulation-institutionalization of this figure in it. Having described the model, its reliability is tested by applying it as a methodology for characterization and evaluation of two case studies, one successful, the Baix Llobregat Agrarian Park, and one frustrated, the proposed Agrarian Park de la Vega of Granada.