🔢 mathematics

Reconstruction of the non-linear wave at a buoy from shoreline data and applications to the tsunami inverse problem for piece-wise sloping bathymetry

This paper demonstrates that initial tsunami wave conditions can be reconstructed from shoreline run-up data on piece-wise sloping bathymetry by recovering boundary conditions at a virtual buoy and integrating non-linear shallow water equations with Boussinesq models to account for dispersion.

Original authors: Oleksandr Bobrovnikov, Madison Jones, Shriya Prasanna, Josiah Smith, Alexei Rybkin, Efim Pelinovsky

Published 2026-02-18

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Oleksandr Bobrovnikov, Madison Jones, Shriya Prasanna, Josiah Smith, Alexei Rybkin, Efim Pelinovsky

Original paper licensed under CC BY 4.0 (http://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

Imagine you are standing on a beach, watching the ocean. Suddenly, a massive wave rushes up the sand, soaking your shoes, and then pulls back. You have a stopwatch and a ruler, and you've recorded exactly how high the water went and how long it took to reach its peak.

Now, imagine you are a detective. Your goal is to figure out what the wave looked like before it hit the shore. Was it a single giant hump? A series of smaller bumps? Did it start far out at sea, or was it generated by a landslide nearby?

This paper is about solving that exact mystery, but with math. The authors have developed a "time-reversal machine" that takes the messy, chaotic data from the shoreline and works backward to reconstruct the wave's original shape and speed, even before it hit the beach.

Here is a breakdown of their work using simple analogies:

1. The Problem: The "Black Box" of the Ocean

Usually, scientists try to predict the future: "If a tsunami starts here, where will it go?" That's the forward problem. It's like throwing a ball and watching where it lands.

But this paper tackles the inverse problem: "We saw where the ball landed; can you tell us how hard and at what angle it was thrown?"

The Challenge: When a tsunami hits the shore, it changes. It slows down, gets taller, and squashes together (non-linear effects). It's like trying to un-mix a smoothie to find out what the original fruit looked like.
The Goal: The authors want to take the "smoothie" (shoreline data) and perfectly reconstruct the "fruit" (the wave's initial shape) so we can know where to place warning buoys in the deep ocean.

2. The Magic Trick: The "Carrier-Greenspan Transform"

To solve this, the authors use a mathematical tool called the Carrier-Greenspan transform.

The Analogy: Imagine the ocean wave is a tangled ball of yarn. It's a mess. The math they use is like a magical needle that instantly untangles the yarn, turning a complicated, twisting knot into a straight, simple line.
What it does: It takes the messy, non-linear physics of a crashing wave and turns it into a simple, linear equation that is much easier to solve backward. It's like turning a complex riddle into a simple math problem.

3. The Two-Step Detective Work

The paper describes a two-part strategy to solve the mystery:

Part A: The Sloping Beach (The "Slide")
Most tsunamis hit a beach that slopes down into the ocean.

The Method: The authors show that if you know how the water runs up the beach (the "run-up"), you can mathematically reverse the process to find out exactly how the water was moving at a specific point offshore (a "virtual buoy").
The Result: They created an algorithm. You feed it the shoreline data, and it spits out the wave's speed and height at the buoy. They tested this with computer simulations and found it works perfectly, even with different shapes of waves.

Part B: The Deep Ocean (The "Flat Floor")
Once they know what the wave looked like at the edge of the shallow slope, they need to figure out what it looked like in the deep, flat ocean.

The Problem: In deep water, waves behave differently. They spread out and have "dispersion" (like a prism splitting light). The simple math used for the beach doesn't work here.
The Solution: They "stitch" two different math models together.
- Model 1 (The Beach): Uses the Non-Linear Shallow Water Equations (good for the slide).
- Model 2 (The Deep): Uses the Boussinesq Equation (good for deep water and spreading waves).
The Analogy: Imagine you are tracking a car. On the highway (deep water), it drives fast and straight. When it hits a muddy field (the beach), it slows down and bounces. The authors figured out how to translate the car's muddy tracks back into its highway speed and direction. They used "solitons" (special, stable wave shapes that don't break apart) as test cases to prove their stitching method works.

4. Why Does This Matter?

You might ask, "Why do we need to know the wave's shape at a buoy?"

Better Warning Systems: Currently, we place buoys (like the DART system) based on guesswork or general rules. This research gives scientists a mathematical way to say, "If we want to detect a tsunami 30 minutes before it hits, we must place the buoy exactly here."
Understanding the Source: By reconstructing the wave, we can better understand the earthquake or landslide that caused it.
Speed: The authors note that their method is incredibly fast. Because they simplified the math, a computer can solve this "time-reversal" problem in seconds, which is crucial for real-time disaster management.

Summary

Think of this paper as a mathematical time machine.

Input: You give it the messy, chaotic record of a wave hitting the beach.
Process: It uses a "magic needle" (Carrier-Greenspan transform) to untangle the physics, then stitches together two different rulebooks (Shallow Water and Boussinesq) to trace the wave back through the deep ocean.
Output: It tells you exactly what the wave looked like when it was first born, allowing us to place our sensors in the perfect spot to catch the next one.

It turns the chaotic roar of a tsunami into a clear, readable story of where it came from.

1. Problem Statement

The paper addresses a specific inverse problem in tsunami modeling: Given the run-up data (vertical oscillations) of a tsunami wave at the shoreline, can one recover the initial shape and velocity of the wave at a virtual buoy located offshore?

Context: Tsunami warning systems (like DART buoys) rely on offshore data to predict coastal inundation. However, buoy placement is critical. This work aims to provide a theoretical framework to determine optimal buoy locations by reconstructing the incoming wave field from historical or synthetic shoreline data.
Challenge: Tsunami propagation near the shore involves strong non-linear effects and complex bathymetry (seafloor topography). Standard linear models often fail to capture these dynamics accurately, while full non-linear models are difficult to invert analytically.
Scope: The study focuses on:
1. Recovering boundary conditions (water displacement $\eta$ and velocity $u$ ) at a finite distance $x=L$ from the shore using Non-Linear Shallow Water Equations (NSWE).
2. Extending this to piece-wise sloping bathymetry, where a sloping near-shore region is stitched to a flat, deep-ocean region modeled by dispersive equations (Boussinesq or Linear Shallow Water).

2. Methodology

The authors employ a multi-stage analytical and numerical approach:

A. The Finite Sloping Region (NSWE)

For the near-shore region ( $0 \le x \le L$ ), the authors use the Non-Linear Shallow Water Equations (NSWE):
$\partial_t \eta + \partial_x(u(x+\eta)) = 0, \quad \partial_t u + u\partial_x u + \partial_x \eta = 0$

Carrier-Greenspan Transform (CGT): To solve the inverse problem analytically, the authors utilize the CGT. This transformation linearizes the non-linear NSWE system and fixes the moving shoreline boundary.
- New variables: $\sigma = x + \eta$ , $\tau = t - u$ .
- Transformed system: A linear wave equation in the hodograph plane $(\sigma, \tau)$ .
Inverse Algorithm:
1. Transform Run-up: Convert observed shoreline run-up $R(t)$ into the hodograph variable $\psi_{sh}(\tau)$ using the CGT relations.
2. Solve Inverse Problem: Apply the Laplace transform to the relationship between the shoreline solution and the boundary solution at $x=L$ . By inverting the Laplace transform, they derive an explicit formula to recover the boundary function $\psi_b(\tau)$ (representing the wave at the buoy) from the shoreline data.
3. Recover Physical Variables: Use the recovered $\psi_b(\tau)$ and the CGT to reconstruct the physical boundary conditions $u(L, t)$ and $\eta(L, t)$ .

B. Piece-wise Sloping Bathymetry (Stitching Models)

To handle the transition from deep water to the sloping beach, the authors propose "stitching" two different models:

Region 1 (Sloping): Governed by NSWE (non-linear, non-dispersive).
Region 2 (Flat/Deep): Governed by either the Boussinesq equation (non-linear, dispersive) or Linear Shallow Water Equations (LSWE).

The Stitching Strategy:

Heuristic Approach: Use the inverse algorithm from Section 2 to recover the wave data at the interface ( $x=L$ ) from shoreline data.
Model Fitting: Fit this recovered interface data to a known multi-soliton solution of the Boussinesq equation (or a traveling wave solution for LSWE).
Initial Condition Recovery: Once the parameters of the solitons (amplitude, speed, position) are extracted from the interface data, the authors work backward to determine the initial conditions of the tsunami in the deep ocean (Region 2) at the time of the generation event.

3. Key Contributions

Analytical Inversion Algorithm: The paper provides a complete, closed-form analytical solution to the inverse problem for NSWE on a finite interval. It demonstrates that boundary conditions (displacement and velocity) at a virtual buoy can be uniquely recovered from a single shoreline run-up function $R(t)$ , provided compatibility conditions are met.
Extension to Piece-wise Bathymetry: The authors successfully bridge the gap between non-linear shallow water models and dispersive models. They demonstrate a method to reconstruct deep-ocean initial conditions (including dispersive effects) using only shoreline data, effectively solving a "hybrid" inverse problem.
Validation via Numerical Experiments:
- Forward-Backward Verification: They manufactured synthetic initial conditions (various pulse shapes), simulated the forward run-up, and then applied their inverse algorithm. The recovered boundary conditions matched the "exact" inputs with high precision.
- Soliton Reconstruction: In the piece-wise model, they successfully recovered the parameters of two interacting solitons from the shoreline run-up, accurately predicting their initial positions and speeds.
Computational Efficiency: The algorithm is computationally efficient ( $O(N^3)$ complexity, reducible to $O(N \log N)$ with FFT assumptions), making it suitable for rapid analysis of tsunami characteristics.

4. Results

Accuracy: Numerical verification showed "excellent agreement" between the exact boundary conditions and those recovered by the algorithm for various initial wave profiles (sech, Gaussian, and composite shapes).
Piece-wise Performance: The stitching method successfully isolated the "incident" wave from the "reflected" wave at the interface. By cropping the time series to remove reflection artifacts, the authors accurately recovered the parameters of a two-soliton solution.
Comparison of Models: The study compared the Boussinesq model against the Linear Shallow Water model for the deep region. Both allowed for the recovery of initial conditions, though the Boussinesq model captures dispersion effects crucial for long-distance propagation.
Physical Insight: The results confirm that even with non-linear effects dominating near the shore, the wave information is preserved in the run-up data and can be mathematically traced back to the offshore source.

5. Significance and Applications

Optimizing Buoy Placement: The primary application is informing the placement of DART buoys. By understanding how well shoreline data can reconstruct offshore waves, authorities can determine the minimum distance required for a buoy to accurately capture the wave's initial shape before non-linear distortion becomes too severe.
Tsunami Warning Systems: The method offers a theoretical basis for "data assimilation" where real-time shoreline data (from tide gauges or video analysis) could be used to retroactively estimate the incoming wave's energy and shape, potentially improving warning times.
Handling Dispersion: By incorporating the Boussinesq equation, the model addresses the debate on whether dispersion matters for tsunamis. The paper shows that dispersion can be accounted for in the inverse problem, allowing for more accurate reconstruction of waves that have traveled long distances.
Future Directions: The authors identify future work in handling multi-sloping bathymetries, defining the set of "admissible" shoreline data (to ensure non-breaking waves), and investigating resonance phenomena that might amplify run-up.

Conclusion

This paper establishes a robust mathematical framework for solving the tsunami inverse problem. It proves that it is possible to reconstruct the complex, non-linear wave field at an offshore buoy solely from shoreline inundation records. By extending this to piece-wise bathymetry and dispersive models, the authors provide a versatile tool for understanding tsunami generation and propagation, with direct implications for improving coastal safety and warning infrastructure.