Saturday, November 21, 2026 The gateway for independent voices — journalism, podcasts & documentaries
RSS
Established for Independent Voices
IJNN.World
Independent Journalist News Network


Breaking
— ‘Fun to see’ former Timberwolf Karl-Anthony Towns in playoffs with Knicks: ex-coach — Columbus Crew visit Inter Miami in first round of MLS Cup Playoffs — james turrell fills new york gallery with floating cubes and glowing walls of light — Magna International Targets 15% EPS Growth as Automation and Buybacks Drive Strategy — ‘Fun to see’ former Timberwolf Karl-Anthony Towns in playoffs with Knicks: ex-coach — Columbus Crew visit Inter Miami in first round of MLS Cup Playoffs — james turrell fills new york gallery with floating cubes and glowing walls of light — Magna International Targets 15% EPS Growth as Automation and Buybacks Drive Strategy

Uncategorized

Degenerate Richards Flow with Signorini Seepage Existence and L1 Contraction

We establish existence and L 1 contraction for a closed, dry-state constrained formulation of the degenerate Richards equation with mixed Dirichlet, Neumann, and Signorini seepage conditions. In the Kirchhoff variable, the seepage boundary becomes a closed convex constraint. An implicit elliptic resolvent contracts the water content in L 1 ; its semigroup limit yields an […]

Degenerate Richards Flow with Signorini Seepage Existence and L1 Contraction

We establish existence and L 1 contraction for a closed, dry-state constrained formulation of the degenerate Richards equation with mixed Dirichlet, Neumann, and Signorini seepage conditions. In the Kirchhoff variable, the seepage boundary becomes a closed convex constraint. An implicit elliptic resolvent contracts the water content in L 1 ; its semigroup limit yields an energy solution with strongly continuous storage. The storage, Kirchhoff state, and Darcy flux are unique within the constructed solution class. Initial water content may be prescribed as an admissible L 1 function without compatibility with the Dirichlet trace at time zero. The argument allows conductivity to vanish at the dry endpoint and uses a quantitative modulus for the transformed gravity flux. This modulus is verified for rational Haverkamp-Vachaud laws and Van Genuchten-Mualem laws with n > 3 / 2 .

The Richards equation is a nonlinear partial differential equation describing water flow in variably saturated porous media under the combined effects of capillarity and gravity [1]. It is obtained by coupling mass conservation with Darcy’s law and by introducing constitutive relationships between the pressure head, the water content, and the hydraulic conductivity. Owing to its ability to describe infiltration, drainage, capillary rise, and groundwater recharge processes, the Richards equation plays a central role in hydrology, soil physics, agriculture, and environmental engineering.

A major difficulty in the analysis and simulation of Richards-type models comes from the strong nonlinearity of the hydraulic functions. In dry regions, the hydraulic conductivity may become very small, or even vanish in idealized models, while the water capacity may also degenerate. As a consequence, the equation may lose its uniformly parabolic character and behave as a degenerate nonlinear parabolic problem. This degeneracy is particularly challenging in the presence of sharp wetting fronts and transitions between saturated and unsaturated zones.

The mathematical analysis of the Richards equation has been extensively studied. Foundational results on nonlinear elliptic-parabolic problems and porous-media type flows were established by Alt and Luckhaus [2] [3], while uniqueness and contraction principles for saturated-unsaturated flows were developed by Otto and Carrillo [4] [5]. Further contributions have addressed degenerate pressure-capillarity relations [6], global existence for subsurface flow models [7] [8], traveling fronts [9], singular-degenerate free boundary problems [10], and related degenerate parabolic equations [11] [12]. These works provide the theoretical background for weak and renormalized formulations of Richards-type equations.

The degenerate Kirchhoff transformation is natural for studying this problem: it controls the capillary flux even when the pressure gradient cannot be bounded in the dry regime. The gravity flux and the unilateral seepage boundary nevertheless require a comparison argument compatible with the closed constitutive graph. Our analysis constructs a contractive elliptic resolvent and then passes to continuous time.

Seepage boundary conditions are especially important in variably saturated flows. They arise when part of the boundary is in contact with the atmosphere and may switch dynamically between an impermeable state and an outflow state. The location of the seepage face is not known in advance and must be determined as part of the solution. This free-boundary character makes the problem more complex than standard Dirichlet or Neumann boundary-value problems, in the analysis. In particular, the active part of the seepage boundary is determined by the solution rather than prescribed in advance.

Earlier contraction and uniqueness results provide the starting point for the present analysis. Otto [4] treats saturated-unsaturated flow with mixed boundary conditions describing impermeable boundaries, reservoirs and contact with the atmosphere. His variational framework allows time-dependent boundary constraints and establishes contraction for finite-energy solutions without assuming an integrable time derivative of the saturation. Carrillo and Wittbold [5] establish comparison and uniqueness for renormalized solutions of degenerate elliptic-parabolic problems with homogeneous Dirichlet boundary conditions, under their structural assumptions on the flux.

In the present work, we develop a direct implicit-resolvent construction for degenerate Richards flow with a Signorini seepage boundary. The analysis uses the Kirchhoff variable , whose gradient remains controlled as the conductivity approaches zero. In this variable, the atmospheric condition takes the form of the unilateral trace constraint on the seepage boundary, while a closed constitutive graph accommodates the saturated plateau. The additional constraint , where is the dry endpoint of the Kirchhoff range, completes the admissible states at residual storage. We consider a fixed boundary partition, with a Dirichlet portion of positive surface measure and stationary Dirichlet and Neumann data. The initial storage may be any element of the closed admissible set , without compatibility with the Dirichlet trace at time zero.

Within this setting, we first establish contraction for the elliptic resolvent and then obtain a strongly continuous contraction semigroup by implicit time discretization. The construction yields existence and uniqueness of the storage, Kirchhoff state and Darcy flux in the associated Rothe energy class. This specifies the solution class covered by our result and its relation to the finite-energy and renormalized frameworks cited above. We also verify the quantitative continuity condition on the transformed gravity flux for rational Haverkamp-Vachaud laws and Van Genuchten-Mualem laws with , thereby identifying constitutive models to which the construction applies.

The paper is organized as follows. Section 0 introduces the Richards model and hydraulic laws, and Section 1 describes the boundary conditions. Section 1 presents the Kirchhoff-Signorini formulation and the main theorem, whose proof by implicit resolvents is given in Section. Finally, Section verifies the constitutive continuity condition and discusses its physical scope.

Water flow in unsaturated porous media is commonly described by the Richards equation. This model is obtained by coupling the mass conservation equation

where [-] denotes the volumetric water content and the macroscopic water flux, with the generalized Darcy law

where [LT−1] is the hydraulic conductivity tensor and [L] the total hydraulic head, defined by

Here, [L] denotes the water pressure head relative to atmospheric pressure, while [L] is the vertical coordinate. The water table is the atmospheric-pressure surface . Positive pressure heads occur in saturated regions, but saturation does not require a positive pressure head: a capillary fringe above the water table may remain saturated under negative water pressure. The water content and hydraulic conductivity are related to through the chosen hydraulic constitutive laws.

Substituting (2) into (1) leads to the classical formulation of the Richards equation:

This equation is supplemented with the initial condition

where Ω is a bounded domain of , with . The boundary is decomposed as

on which Dirichlet, Neumann, and seepage boundary conditions are prescribed on , , and , respectively.

The formulation (4) involves coefficients that depend nonlinearly on the matric potential. It is therefore necessary to specify the hydraulic relationships linking the water content and the hydraulic conductivity to the pressure head.

Hydraulic properties constitute a central element in the modeling of the Richards equation, as they describe how an unsaturated soil stores and transmits water as a function of the matric potential. The soil-water retention relationship expresses the volumetric water content , which is generally monotone increasing and bounded between the residual water content and the saturated water content [3] [4]. It is often reformulated in terms of the effective saturation

which normalizes the water content between 0 and 1.

The derivative , called the specific moisture capacity, measures the sensitivity of the water content to pressure variations. It appears in the time-derivative term of the Richards equation and may become very small in dry regions, leading to a parabolic degeneracy of the model.

The air-entry pressure head , when the retention model has a distinct entry threshold, marks the onset of desaturation as the pressure head decreases. It must be distinguished from the water table, which is defined by . For , the interval may describe capillary-saturated states above the water table; desaturation occurs below the air-entry threshold. In a hydrostatic profile, , where is the water-table elevation, so the idealized saturated capillary fringe extends from to . Some empirical laws approach saturation continuously without a separate negative air-entry plateau. In particular, the standard examples used below are extended by saturation at ; this is a constitutive idealization and does not identify atmospheric pressure with a universal physical air-entry threshold. A piecewise law with a specified threshold can be written as follows:

where and denote the constitutive laws adopted in the unsaturated zone.

The hydraulic conductivity, which also depends on the matric potential, is classically written as

where denotes the saturated hydraulic conductivity. In the unsaturated zone, decreases sharply and may reach extremely small values, thereby enhancing both the nonlinearity and the degeneracy of the equation.

The functions , , and do not arise from an analytical model but from empirical laws fitted to experimental data. The most commonly used ones are the Gardner-Irmay, Vachaud, and, in particular, the Van Genuchten-Mualem relationships [13]-[16], which have become a standard for representing soil hydraulic properties. These laws are continuous and monotone in their physical ranges; the additional modulus required for uniqueness is checked in Section, rather than being assumed from continuity alone. They play a crucial role in the mathematical analysis of the Richards equation. The main formulations used in this work are summarized in Table 1.

Table 1. Hydraulic constitutive laws used in the unsaturated zone.

Beyond the definition of the hydraulic properties of the medium, solving the Richards equation requires the specification of appropriate boundary conditions capable of representing the different flow regimes.

The solution of the Richards equation relies crucially on the choice of appropriate boundary conditions, as they govern the interaction between the porous medium and its environment and influence the transition between saturated regimes (elliptic in nature) and unsaturated regimes (degenerate parabolic in nature). The three main families of boundary conditions are Dirichlet, Neumann, and the so-called seepage condition, which is specific to free-surface flows in unsaturated porous media.

On a portion of the boundary , the pressure head is prescribed directly as

where may depend on space and time. This condition is typically used to model an imposed water level, a groundwater table, or a reservoir.

On another part of the boundary, the normal water flux is imposed. Denoting by the total hydraulic head and by

where is the outward unit normal vector. Thus denotes exfiltration, denotes infiltration, and corresponds to an impermeable boundary.

On an exposed soil boundary, the water pressure may evolve between negative values, for which the idealized seepage condition prevents outflow, and atmospheric pressure, at which outflow is allowed. Prescribing either a pressure or a flux in advance does not by itself describe this switching behavior. This phenomenon leads naturally to the seepage condition, which is essential for modeling processes such as surface exfiltration, natural drainage of embankments, water outflow along riverbanks, or localized seepage in hydraulic structures [18] [19].

3.4. Physical and Mathematical Formulation of the Seepage Condition

For an exposed boundary in contact with the atmosphere, the idealized seepage condition permits outflow only when the water pressure reaches atmospheric pressure. If , the condition imposes zero outward flux; this branch is characterized by negative pressure, rather than by an assertion that the adjacent soil must be unsaturated. If , nonnegative outward flux is permitted. These alternatives are expressed by the following complementarity system:

The active outflow subset of the prescribed seepage boundary is unknown in advance. Its boundary separates active and inactive seepage portions; it should not be identified in general with the entire interface between saturated and unsaturated soil.

At each boundary point, the idealized seepage condition has two branches: with zero outward flux, and with nonnegative outward flux. The analytical convex set defined in Section 1 imposes the corresponding unilateral trace constraint in the Kirchhoff variable, together with a lower dry-state bound. Recovery of the classical flux complementarity from the variational inequality requires identifiable conormal traces and the absence of an active dry reaction at the boundary, as explained in Section 1.

In this work, the boundary may simultaneously include Dirichlet, Neumann, and seepage portions. The outflow region is part of the solution rather than part of the prescribed boundary partition. Figure 1 illustrates the atmospheric-pressure water table and the active seepage portion on an inclined wall; a capillary-saturated fringe is not resolved in this schematic.

Figure 1. Schematic atmospheric-pressure water table and active seepage boundary on an inclined wall. The surface defines the water table. Negative pressure above it does not exclude a saturated capillary fringe, which is not resolved in this schematic. The active seepage segment is determined by the Signorini condition and is not a prescribed Dirichlet boundary segment.

The previous modeling elements now allow us to rigorously formulate the variational problem associated with the Richards equation and to state the main result of this work.

For the analysis we factor the conductivity tensor as , where satisfies

Let be positive at every finite pressure, bounded, and constant for ; let be bounded and non-decreasing, and constant for . We assume that is integrable at the dry end and write

The functions are extended continuously to , with and constant on . Denote the endpoints of the storage range by and . Set

The dry endpoint represents the limiting residual-water state, rather than a finite pressure head. The transformed Darcy flux is

Let be a bounded Lipschitz domain whose boundary is partitioned into , , and , with . Throughout the analytical theorem, the Dirichlet lifting and the outward Neumann flux are independent of time, and .

We use the space of test functions

We assume . The lower bound completes the pressure range at its dry endpoint, where the pressure head has no finite value. It is an obstacle constraint, and the theorem concerns the resulting closed-graph problem, including its normal-cone reaction. Because the constraint is imposed on states, it also restricts their boundary traces. Consequently, at dry contact the variational formulation must not automatically be identified with the source-free Richards equation together with every prescribed boundary flux realized as an equality. Where the solution is sufficiently regular and separated from , two-sided admissible variations recover the corresponding unconstrained balance and boundary conditions.

Since , the atmospheric condition is exactly on . For regular states separated from the dry endpoint near the boundary, the variational inequality has the form

It yields , , and on the seepage boundary where the conormal trace exists and the dry constraint is inactive.

For sufficiently regular states admitting a multiplier representation, the lower obstacle can be described in the interior by

On a Neumann boundary portion, its formal boundary counterpart is

Thus, an outward extraction demand need not remain fully realized once the residual-water constraint is active. The term represents the unrealized part of that boundary demand in this multiplier description. If a bulk reaction is present, it contributes to the balance; it cannot be omitted while claiming a source-free equation. A dry reaction can also affect the recovery of the classical seepage trace conditions at a dry boundary point.

When the indicated quantities and normal traces admit integration, the balance associated with these relations reads

In particular, the usual balance with the prescribed Neumann flux and no volume source requires the absence of these dry-reaction terms. These formulas explain the constraint at a regular or formal level. The general theorem does not assert that its reactions have integrable densities or that all the displayed normal traces exist; its unconditional statement remains the integrated variational formulation.

A simple example illustrates the distinction between requested and realized extraction. Take a nonnegative, nonzero Neumann datum , the constant lifting , and initial storage . The state , satisfies every elliptic step: the storage increment and the interior flux vanish, while

Here , as follows from dry-end integrability and the continuous extension of . The actual Darcy flux is therefore zero, although the requested outward flux is positive. This example belongs to the closed dry-endpoint formulation; the datum represents a limiting dry pressure, not a finite imposed pressure head.

The constrained formulation represents the original source-free Richards problem with its stated boundary conditions when the dry normal-cone contribution vanishes and the balance and boundary fluxes can be identified in the relevant weak or classical sense. A sufficient setting for the usual formal identification is a sufficiently regular solution uniformly separated from in the region and up to the boundary portions under consideration. Strict separation is a sufficient condition, not a necessary one: contact alone does not imply a nonzero reaction. No unconditional absence of dry reaction is claimed for arbitrary admissible initial storage and extraction data.

The two structural assumptions used below are

Here is non-decreasing and tends to zero at zero. Compatibility means for a continuous function . The saturated plateau of the Haverkamp and Van Genuchten-Mualem laws satisfies this condition.

For a precise weak formulation, when choose a non-decreasing generalized inverse of the storage law and set

When the storage range is a singleton, set . The function is convex and Lipschitz. With the subdifferential taken relative to ,

At an endpoint, this subdifferential also includes the outward normal cone of the storage interval; equality with the inverse storage graph is therefore not asserted. In particular, (16) includes the saturated plateau and the dry endpoint . Write

Definition 1 (Rothe energy solution). A Rothe energy solution with initial storage is a limit, as the uniform step , of the fully implicit scheme (22) initialized by . It has

No trace compatibility between and the stationary Dirichlet datum is required; the Dirichlet trace is imposed on for almost every . The integrated variational inequality is, for every and ,

The pairing on the right is well defined for , since the storage is bounded. This formulation records the Signorini normal cone without postulating a separate bound on .

Theorem 1 (Existence and L1 contraction). Under the geometric, data, and constitutive assumptions above, for every the degenerate Kirchhoff-Signorini problem has a Rothe energy solution with initial storage . For two such solutions with initial storages and the same boundary data, their storages satisfy

The Kirchhoff state and the flux (11) are uniquely determined by the initial storage within this Rothe energy class. The construction uses the closed degenerate constitutive graph itself; no positive lower bound on or uniform estimate of is required.

5. Implicit Construction and Proof of the Main Result

Let , let , and seek such that and

Although the storage range is bounded, need not be bounded; the integral in (19) is nevertheless defined since and .

Proposition 2 (Resolvent and contraction). Problem (19) has a unique for each and . The map satisfies

Proof. The operator in (19) is pseudomonotone on the closed convex set . The Nemytskii storage operator is monotone and demicontinuous. If in , Rellich compactness gives convergence in measure, and bounded continuity of gives strongly in . Thus the gravity term is a strongly continuous perturbation from into . Taking a fixed , ellipticity, Poincaré’s inequality, and the trace theorem give

For contraction put and . The functions and are pointwise between the states, so both are admissible. The Neumann terms cancel. Adding the two inequalities and dividing by yields

On , ellipticity and Young’s inequality bound the right-hand side, after absorbing half of the diffusion term, by

This tends to zero by (15). Monotonicity of gives ; the equality also holds across storage plateaus because there. Letting at the elliptic step gives

Interchanging indices gives the full contraction. For , the storages coincide. Compatibility (14) then makes the two gravity terms equal. Testing the two original inequalities with and , respectively, proves . The common Dirichlet trace implies . □

Let , set , and for solve (19) recursively with

Proposition 2 proves existence and uniqueness at every step without a step-size restriction. Fix , independent of time, and define

The subgradient relation (16) implies . It is finite for every , because and are bounded and . Since ,

Testing the -th resolvent by gives

Put . Ellipticity, boundedness of , the trace theorem, and Poincaré’s inequality give constants , independent of , such that

Summation of (25) and nonnegativity of the relative energy prove

Consequently the piecewise constant interpolants are uniformly bounded in and , respectively. Neither nor any pressure preimage of it needs an trace compatible with the Dirichlet datum.

We now obtain temporal compactness from the resolvents. The family is nonexpansive by (20). Its resolvent identity follows directly by writing in (19): for ,

For , , testing the resolvent by and using coercivity of gives a bound on in , independent of , and

For each , uniform continuity of on (it is constant on ) gives such that implies . Monotonicity therefore yields on this set. The preceding estimate implies

and hence in . Moreover is dense in for the norm. Indeed, approximate any by simple storages, choose their preimages in , approximate the resulting bounded states in the interior by smooth functions, and join them to a fixed in a boundary layer of vanishing measure. Convex interpolation preserves the lower bound , while the fixed boundary trace and seepage constraint come from . Boundedness of makes the layer error vanish in . Contraction extends to every .

Lemma 3 (Accretive realization of the resolvents). Let and define the graph by

Then is accretive in , and

Moreover, for every , and these resolvents leave invariant.

Proof. Take , , and choose representations and , with and . If , convexity of gives

The resolvent identity yields , and contraction gives

To obtain the same inequality for an arbitrary , choose . By convexity of the norm,

For and , the defining graph contains , so . Accretivity makes the inverse single-valued on its range; hence its value at is exactly . The invariance follows from . Finally, , and is closed in . Conversely, for every , the already established convergence in as , with , gives . This proves the asserted equality. □

We apply the following form of the Crandall-Liggett generation theorem [20]: if is an accretive operator on a Banach space and for all sufficiently small , the exponential formula generates a strongly continuous contraction semigroup on . For the graph above, the closure of the domain is , the required range inclusion holds for every , and the resolvents on are precisely the maps already constructed by the variational inequality. Thus

with , defines the desired semigroup on . The associated implicit-Euler convergence, uniform on bounded time intervals, yields the following convergence for the Rothe interpolants:

The harmless one-step convention at is fixed by . In particular, temporal translations of the storage vanish uniformly along the scheme. This is the temporal compactness needed for the degenerate problem; no estimate of a separate Signorini reaction is invoked.

The energy bound yields, along a subsequence, in . Equation (29) gives in every finite , since the storages are uniformly bounded. The monotone-graph inequality

for bounded , followed by weak-strong passage and Minty’s argument, identifies . Convex weak closure gives almost everywhere. Compatibility gives

and therefore the fluxes converge weakly in to (11).

For completeness, the limiting Signorini inequality can be obtained without assigning an derivative to the storage. Convexity of implies

Sum (19) between two time levels with a fixed , apply this inequality, and pass to the limit. The storage terms converge strongly, including at the two endpoints of the interval, by (29). The negative quadratic form is weakly upper semicontinuous; all linear flux terms pass weakly, and the gravity coefficient passes strongly. We obtain precisely (17). Piecewise constant admissible competitors yield the corresponding time-dependent integrated formulation. This proves existence in Theorem 1.

Apply (20) at successive time levels and pass to (29) to obtain (18). In particular, equal initial storages generate the same continuous storage trajectory. The next lemma proves that this trajectory determines the Kirchhoff state and the flux without pairing a distributional time derivative of the storage with a moving competitor.

Lemma 4 (Reconstruction of the Kirchhoff state). Let and satisfy (17) with the same storage , the same set , and the same Neumann datum. Assume almost everywhere, , and (14). Then almost everywhere in , and their Darcy fluxes coincide.

Proof. Put . Choose such that , almost everywhere, and is a Lebesgue point of both . These properties hold for almost every . At this fixed , use the admissible competitor in (17) for , and for , on . Adding the two inequalities gives

Thus, the sum of the two time integrals in (30) is nonnegative. Divide it by and let . At the chosen Lebesgue point, the mean of over tends to zero. Bounded continuity of then gives convergence

of the gravity term, and the fixed boundary datum passes by the trace theorem. Consequently,

Because , compatibility implies . The gravity and Neumann terms cancel, leaving

Ellipticity makes the gradient difference zero. The two states have the same Dirichlet trace on , so Poincaré’s inequality gives . This holds for almost every ; the flux formula (11) then proves the last claim. □

This completes the proof of Theorem 1.

The key order estimate is the exact resolvent inequality (20). The limit is taken before the time-step limit, so no time-chain assertion for an unspecified weak solution is needed. The outward Neumann flux cancels because the data are the same and stationary; the seepage cone is preserved by the lattice competitors in the proof.

Hence is Lipschitz for any and . The Vachaud rational conductivity has the same form and therefore satisfies the same modulus condition when its exponent is greater than one. A corresponding strictly increasing retention law below saturation, with a compatible saturated plateau, fulfills the constitutive assumptions.

With the standard and , the dry conductivity behaves as , . The dry Kirchhoff endpoint is finite and . At saturation and . Therefore

so (15) holds. Both laws have constant storage and conductivity above saturation, which verifies (14). For , the stated modulus hypothesis does not hold; Theorem 1 is asserted for the specified range.

The condition is a sufficient condition for the transformed gravity-flux modulus used in the present contraction proof. Its origin is the behavior near saturation. For the standard choice ,

The resulting Hölder exponent is , so the stated hypothesis holds when . In contrast, the dry-end conductivity satisfies with for every . Thus the restriction is not caused by dry-end integrability. For , the present modulus hypothesis fails; this does not prove nonexistence or nonuniqueness for the original constitutive law.

The restriction excludes physically relevant parameterizations of fine-textured soils. Table 2 lists representative mean values from Carsel and Parrish [22]. Several clay-rich classes have values close to one, and the exclusion also affects some silt-rich classes. These catalog values illustrate the scope of the theorem; they are not universal constants for a texture class. Applicability to a particular soil must be assessed using its fitted hydraulic parameters. Moreover, satisfying this exponent condition does not dispense with the other constitutive, geometric and boundary assumptions of the theorem.

Table 2. Illustrative mean Van Genuchten exponents from Carsel and Parrish [22] and satisfaction of the exponent condition in the present theorem.

One possible extension is to regularize the conductivity near saturation. For a fixed , one may retain the original law for , replace it on by a positive, monotone Lipschitz connection to , and keep the constant saturated extension. The Kirchhoff transformation must then be recomputed for the modified law. Since the conductivity is bounded away from zero on this transition interval, a bounded derivative there gives a bounded ratio and hence a Lipschitz transformed conductivity on that interval. Together with the unchanged regular dry-end behavior and the compatible storage plateau, this provides a route to a model satisfying the present assumptions for each fixed regularization. Regularizing only the dry tail would not remove the near-saturation obstruction. The relevant constants can deteriorate as , so contraction and uniqueness for the unregularized law cannot be inferred without an additional limiting argument.

An alternative is to fit a constitutive pair with a compatible near-saturation behavior and a transformed conductivity satisfying the required modulus. The rational Haverkamp-Vachaud conductivity with exponent , already examined above, is one such mathematical possibility. Replacing a constitutive law changes the physical model and must therefore be assessed against the hydraulic data of the soil under consideration. Extending the contraction argument to the unmodified Van Genuchten-Mualem law throughout remains outside the present theorem.

We have constructed a Rothe energy solution for the closed dry-state constrained Kirchhoff formulation of degenerate Richards flow with stationary Dirichlet and Neumann data and a Signorini seepage boundary. The initial storage can be any admissible state; its pressure need not satisfy the Dirichlet trace at time zero. The elliptic resolvent is contractive in the storage, and its semigroup limit is strongly continuous in time. The resulting storage, Kirchhoff state, and Darcy flux are unique in the constructed energy class under the stated gravity modulus. Rational Haverkamp-Vachaud conductivities and Van Genuchten-Mualem laws in the specified exponent range satisfy this modulus.

A modeling limitation is the spatially uniform retention law and the separable conductivity . The measurable uniformly elliptic tensor permits spatial variations and jumps in conductivity magnitude and anisotropy, but the pressure dependence of the scalar conductivity and the retention curve are common to the medium. Consequently, the theorem does not directly cover general layered soils whose layers possess independently varying retention and relative-conductivity curves, including different residual and saturated water contents or different Van Genuchten parameters. Such extensions require a treatment of spatially dependent constitutive graphs and appropriate interface transmission conditions; they do not follow by simply replacing in the present proof. At dry contact, the source-free interpretation is subject to the qualifications stated in Section 1.

Extension to time-dependent boundary data requires a corresponding treatment of the moving admissible set.

This work was supported by the ADEN-MED project (Adaptability to Extreme Events and Natural Risks-Application to the Mediterranean and Djibouti), funded by the Région Sud Provence-Alpes-Côte d’Azur under the AAP MEDCLIMAT “Natural Risks and Food Sovereignty”; the France 2030 Priority Research Program and Equipment (PEPR) Maths-Vives-Mathematics in Interactions, targeted project HYDRAUMATH (ANR-23-EXMA-007), operated by the French National Research Agency (ANR); the ANTARES project, funded by the ANR; and the PHAROS project (Monitoring Hydrodynamic Processes for Improved Adaptability to Oceanographic and Submergence Risks), funded by the Région Sud Provence-Alpes-Côte d’Azur under the AAP Océan et Mer.…Read more by Yacouba Zongo, Mehmet Ersoy, Alioune Ba, Scientific Research Publishing, Alioune Ba;Mehmet Ersoy;Yacouba Zongo

‹ Tosca's 2026 Sustainability Report Makes the… UBS sees Deutsche Boerse shares overvalued… ›