Abstract:We construct Hölder continuous, global-in-time probabilistically strong solutions to 3D Euler equations perturbed by Stratonovich transport noise. Kinetic energy of the solutions can be prescribed a priori up to a stopping time, that can be chosen arbitrarily large with high probability. We also prove that there exist infinitely many Hölder continuous initial conditions leading to non-uniqueness of solutions to the Cauchy problem associated with the system. Our construction relies on a flow transformation reducing the SPDE under investigation to a random PDE, and convex integration techniques introduced in the deterministic setting by De Lellis and Székelyhidi, here adapted to consider the stochastic case. In particular, our novel approach allows to construct probabilistically strong solutions on $[0,\infty)$ directly.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is about the global existence and non - uniqueness of the three - dimensional Euler equations under the perturbation of transport noise. Specifically:
1. **Global Existence**:
- The paper proves that for any given energy distribution \(e(t)\), there exists a global, probabilistically strong and analytically weak solution \((v, q)\) of the three - dimensional Euler equations, whose kinetic energy is equal to \(e(t)\). These solutions have Hölder continuity and can maintain this property in arbitrarily large time intervals.
- The specific result is: for any constants \(\kappa\in(0, 1)\) and \(T\in(0,\infty)\), and for any given energy distribution \(e: [0,\infty)\to(0,\infty)\), there exists a stopping time \(t\) such that \(P\{t\geq T\}\geq\kappa\), and there exists a Hölder - continuous, probabilistically strong and analytically weak solution \((v, q)\) of the three - dimensional Euler equations, such that the kinetic energy satisfies \(\int_{T^3}|v(x,t)|^2\,dx = e(t)\) for all \(t\in[0,t]\).
2. **Non - Uniqueness**:
- The paper also proves that there exist infinitely many Hölder - continuous initial conditions \(u(\cdot,0)\) such that the solutions of the associated Cauchy problems are not unique on the normal.
- The specific result is: for any two energy distributions \(e_1\) and \(e_2\), if they are equal in a certain time interval \([0,T/2]\), that is, \(e_1(t)=e_2(t)\) for all \(t\in[0,T/2]\), then the two global probabilistically strong solutions \((v_1,q_1)\) and \((v_2,q_2)\) given by Theorem 1.3 are the same in this time interval, but have different kinetic energies when \(t\in(T/2,T]\), thus proving the non - uniqueness of the solutions.
### Main Technical Means
To prove the above results, the paper adopts the following main technical means:
1. **Flow Transformation**:
- By introducing a flow \(\varphi(x,t)=x+\sum_{k\in I}\int_0^t\sigma_k(\varphi(x,s))\circ dB_k(s)\) generated by noise, the original stochastic partial differential equation (SPDE) is transformed into a random ordinary differential equation (random PDE).
- This transformation enables the "removal" of the transport term in the equation, thus simplifying the problem handling.
2. **Convex Integration Method**:
- Use the convex integration method to construct a sequence of smooth approximate solutions \((v_n,q_n,\varphi_n,\mathcal{R}_n)\), and construct the final solution by controlling the convergence of these approximate solutions.
- Specifically, gradually reduce the size of Reynolds stress \(\mathcal{R}_n\) through an iterative process and control the Hölder regularity of the solutions.
3. **Introduction of Stopping Time**:
- Introduce a series of stopping times \(t_L\) to control the growth of solutions and ensure the uniform boundedness of solutions in the appropriate Hölder space.
### Conclusion
This paper for the first time proves that in the three - dimensional Euler equations perturbed by transport noise, there exist global, probabilistically strong and analytically weak solutions, and these solutions are non - unique in some cases. This result is of great significance for understanding the stochastic effects in fluid dynamics.