🔬 physics

Analytical solution for the inverse run-up problem

This paper presents an analytical solution for the inverse run-up problem to reconstruct initial tsunami forcing conditions from shoreline time histories, demonstrating that the method effectively handles real-world data challenges such as uneven sampling, noise, and uncertainties, thereby offering a valuable tool for studying historical tsunamis with unknown sources.

Original authors: Cesare Angeli, Ira Didenkulova, Alberto Armigliato

Published 2026-10-08

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Cesare Angeli, Ira Didenkulova, Alberto Armigliato

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

When a massive earthquake shifts the ocean floor, it can displace a vast column of water, sending a long, powerful wave racing toward the shore. As this wave reaches shallow water, it slows down but grows taller, eventually crashing onto the beach in a process scientists call "run-up." For decades, researchers have focused on the forward problem: taking a known earthquake or wave shape and calculating how high the water will climb up the coast. This helps in predicting disasters and designing seawalls. However, the reverse situation is often more difficult and more critical. Imagine a tsunami has already happened, perhaps a historical event where no instruments were present, or a modern one where the sensors failed. All that remains are the scars on the land, the memories of witnesses, or a few scattered measurements of how high the water reached and when. The challenge is to work backward from these limited clues to reconstruct the original wave that caused them. This is the "inverse problem," a puzzle where the goal is to deduce the hidden cause from the visible effect.

A team of researchers has now developed a precise mathematical method to solve this puzzle for long ocean waves on a sloping beach. Their work provides a clear, analytical way to take a record of how the shoreline moved over time and calculate exactly what the water surface looked like far out at sea before the wave arrived. While many scientists rely on complex computer simulations to tackle this, the authors show that a direct, exact formula exists. This approach is particularly valuable because it can handle real-world data that is messy, incomplete, or recorded at irregular intervals, such as the sparse notes left by eyewitnesses or the rough measurements from a field survey after a disaster.

The core of their discovery relies on a specific type of mathematical relationship known as an Abel integral equation. In simple terms, this equation describes how the initial shape of the wave is mathematically linked to the way it runs up the beach. The researchers found that if you know the history of the water's height at the coast, you can reverse this equation to find the original wave shape. However, there is a crucial catch: the physics of a real tsunami involves non-linear effects, meaning the wave behaves differently as it gets taller and steeper. If one tries to use the simple formula directly on a real, non-linear wave, the result is a distorted picture that overestimates the size of the original wave. The authors demonstrate that by first applying a specific correction step—essentially translating the complex, real-world motion into a simpler, linear version—one can use the exact formula to get the right answer. Without this translation step, the reconstructed wave would appear larger and more energetic than it truly was.

To make this method practical for real data, which rarely comes as a perfect, continuous curve, the researchers showed how to break the run-up history into simple straight-line segments. This technique, known as using linear splines, allows the formula to work with just a handful of data points. In their tests, they found that even with very few measurements, the method could accurately reconstruct the original wave. They also examined what happens when the data is noisy, such as when wind waves or small ripples are mixed in with the tsunami signal. Their analysis revealed that while this noise does create errors in the reconstruction, those errors are strongest near the coast and fade away as you look further out to sea. This means that even with imperfect data, the method can still provide a reliable picture of the wave's origin far offshore.

The team tested their approach with a synthetic example designed to mimic a real-world scenario where data is extremely scarce. They imagined a situation where the only information available was the time the water first retreated, the times and heights of the highest and lowest points of the wave, and a single observation of the water level dropping by a specific amount near a landmark. Using a statistical approach that accounts for the uncertainty in these rough measurements, they generated dozens of possible wave histories. When they averaged these results, the reconstructed initial wave matched the true, known wave with remarkable accuracy. The method successfully identified the shape and size of the original disturbance, proving that it is possible to recover the source of a tsunami even when the evidence on the shore is fragmentary and uncertain.

This work offers a powerful new tool for understanding tsunamis, especially those from the past where no digital records exist. By turning the problem of reconstructing a wave into a solvable equation, the researchers have provided a way to look back at historical events and learn from them. The method is not just a theoretical exercise; it is designed to work with the messy, incomplete data that scientists actually have to deal with in the field. Whether analyzing the aftermath of a recent event to help emergency responders or piecing together the story of a centuries-old disaster, this approach allows us to see the invisible wave that started it all, based solely on the marks it left behind on the land.

Technical Summary: Analytical Solution for the Inverse Run-up Problem

Problem Statement
The paper addresses the "inverse run-up problem" for long water waves, specifically tsunamis generated by seismic events. While the "direct" problem (computing shoreline motion from known initial conditions) is well-studied analytically and numerically, the inverse problem—reconstructing the initial water surface displacement ( $\eta_0(x)$ ) from observed run-up time histories ( $\eta_{sh}(t)$ ) at a single coastal location—remains challenging. This is particularly relevant for scenarios where source data is unknown or uncertain, such as historical tsunamis or events with sparse observational data. The study focuses on the specific case of zero initial velocity, a standard assumption for seismic tsunami generation (the "piston model"), on a plane beach with linear bathymetry ( $h(x) = \alpha x$ ).

Methodology
The authors develop a fully analytical solution based on the following steps:

Mathematical Formulation: The problem is governed by the 1D nonlinear shallow water equations. The authors utilize the Carrier-Greenspan transformation to linearize the system. Crucially, they distinguish between the nonlinear shoreline motion and the linearized solution.
Inverse Riemann Transform: To solve the inverse problem, the authors first apply an inverse Riemann transform to the observed run-up time history. This step is essential to linearize the data; applying the subsequent inversion directly to nonlinear data leads to significant biases (overestimation of amplitude and volume).
Abel Integral Equation: Once linearized, the relationship between the initial condition and the run-up history is shown to be an Abel integral equation. The authors derive an explicit analytical inversion formula (Eq. 18) that recovers $\eta_0(x)$ from $\eta_{sh}(t)$ .
Linear Spline Approximation: To handle real-world discrete data and avoid singular integrals, the authors reformulate the solution using a change of variables that removes singularities. They propose approximating the run-up time history as a series of linear splines (piecewise linear functions). This allows for an explicit, non-singular analytical solution (Eq. 25) for each segment, which can be superposed to reconstruct complex time histories.
Noise Analysis: The paper analytically investigates the effect of high-frequency noise (simulating wind waves) on the inversion. By modeling noise as a sinusoidal perturbation, the authors derive an exact analytical expression for the error in the reconstructed initial condition, showing that the noise-induced disturbance decays with distance from the shore following Green's law ( $x^{-1/4}$ ).
Synthetic Case Study: A synthetic example is constructed using sparse, unevenly sampled data points mimicking eyewitness accounts (e.g., timing of extrema and specific draw-down levels). A Monte Carlo-like procedure is employed to generate an ensemble of possible run-up histories based on measurement uncertainties, which are then inverted to assess the robustness of the reconstruction.

Key Contributions and Results

Explicit Analytical Inversion: The paper provides a closed-form analytical solution for the inverse run-up problem under the assumption of zero initial velocity and linear bathymetry. This avoids the computational cost and potential non-uniqueness issues of purely numerical inversion methods.
Necessity of Linearization: The study demonstrates that ignoring the inverse Riemann transform (i.e., treating nonlinear run-up data as linear) introduces a systematic positive bias in the reconstructed initial condition. This bias increases with the breaking parameter ( $Br$ ) and wave nonlinearity.
Efficiency of Linear Splines: The authors show that representing run-up data as linear splines allows for rapid, accurate reconstruction of the initial condition. Numerical tests indicate that even with relatively few sampling points (e.g., $N=20$ ), the spline-based inversion reproduces the original initial condition with high precision.
Noise Robustness: The analytical noise analysis reveals that while high-frequency noise affects the near-field reconstruction, its impact diminishes offshore. The dominant error term decays as $x^{-1/4}$ , suggesting that offshore initial conditions are relatively insensitive to near-shore noise contamination.
Sparse Data Application: The synthetic example demonstrates that the method can successfully reconstruct initial conditions from very sparse data (typical of historical records or eyewitness accounts). By using an ensemble approach to account for measurement uncertainty, the average of the inverted solutions closely approximates the true initial condition, particularly in the far-field.

Significance and Claims
The authors position this work as a tool for gaining "invaluable insight" into the physical processes of tsunami generation and propagation, complementing numerical approaches. The primary significance lies in providing a fast, exact analytical framework for reconstructing tsunami sources from limited coastal observations.

The paper claims that this approach is particularly applicable to:

Historical Tsunami Studies: Reconstructing sources for events where only sparse field survey data or eyewitness testimonies exist.
Rapid Source Estimation: Providing quick, approximate source characterizations immediately after an event using limited real-time run-up observations.

The authors are modest regarding the scope, acknowledging that the solution is limited to 1D linear bathymetry and zero initial velocity. They note that while generalizations to 3D geometries and multilinear bathymetries exist for the direct problem, extending the inverse solution to these complex geometries remains a future challenge. The work serves as a foundational analytical step, offering a benchmark for understanding the inverse problem before tackling more complex, real-world geometries.

Technical Summary: Analytical Solution for the Inverse Run-up Problem

More like this