📄 earth science

A Three-Component Spliced Lognormal-Weibull-K Distribution Framework for Seasonal Modeling of Significant Wave Height in the Abadi-1 Masela Offshore Area

This study proposes a novel three-component spliced Lognormal–Weibull–K distribution framework to accurately model the seasonal variability of significant wave height across low, moderate, and upper-tail regimes in the Abadi-1 Masela offshore area, demonstrating its superior capability for assessing operational risks compared to single-distribution models.

Original authors: Gysberth Maurits Wattimena, Sefnath Johanes Wattimena, Roy Robert Lekatompessy

Published 2026-10-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Gysberth Maurits Wattimena, Sefnath Johanes Wattimena, Roy Robert Lekatompessy

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

The ocean is a place of constant, churning motion, but for the engineers and crews who work on its surface, the most critical number is not the speed of the wind or the depth of the water. It is the height of the waves. Specifically, they track the "significant wave height," a measurement that represents the average height of the largest third of the waves in a given area. This single number dictates whether a ship can safely dock, whether a drilling rig can continue its work, or if a subsea pipeline is at risk of being damaged by the force of the water. For decades, scientists have tried to predict these heights using mathematical models that assume waves follow a single, consistent pattern. However, the ocean rarely behaves with such simplicity. In many locations, the waves that crash against a ship's hull are not all created equal; the small, gentle swells that dominate most days behave differently than the rare, towering walls of water that appear during storms. Treating all these different types of waves as if they belong to the same family can lead to dangerous miscalculations in safety planning and operational scheduling.

In the waters off the eastern coast of Indonesia, near the massive Abadi-1 Masela gas project, a team of researchers has proposed a new way to understand this complexity. They focused on a stretch of the Arafura Sea where the climate is driven by the shifting monsoons, creating four distinct seasons of wave activity. Instead of forcing all the wave data into one single mathematical box, the researchers decided to split the ocean's behavior into three separate zones: the low waves that are common and calm, the moderate waves that are frequent enough to slow down work, and the extreme upper-tail waves that are rare but potentially catastrophic. By treating each of these zones as its own distinct entity, they built a more flexible and accurate picture of the sea.

The study analyzed three-hourly measurements of wave height collected between 2022 and 2026. The team found that the ocean in this region does not follow a uniform rhythm throughout the year. During the East Monsoon, the sea is at its most energetic, with an average wave height of 1.539 meters. In contrast, the period known as Transition II is the calmest, with an average height of just 0.929 meters and no recorded instances of the extreme high waves that define the most dangerous conditions. The West Monsoon presents a different kind of risk; while its average wave height is not the highest, it produces the most frequent occurrences of those rare, towering waves, with the largest recorded height reaching 3.870 meters. This distinction is vital because the danger to offshore operations often comes not from the average conditions, but from those infrequent, extreme spikes.

To capture these nuances, the researchers developed a "spliced" model, which is essentially a patchwork of three different statistical descriptions stitched together. For the low, gentle waves, they used a distribution that fits the behavior of small, frequent swells. For the moderate waves that make up a large portion of the year, they applied a different mathematical shape that better describes the transition zone. Finally, for the rare, high-energy events, they used a third model specifically designed to handle the extreme upper limits of the data. This approach allowed them to see clearly how the ocean behaves in each season. For instance, during the West Monsoon, the model showed that while most days are calm, the risk of extreme waves is real and measurable, accounting for about 4.12 percent of the total data. In the East Monsoon, the sea is dominated by moderate conditions, which are frequent enough to restrict certain types of heavy lifting or drilling operations, even if they are not dangerous enough to stop work entirely.

The researchers compared their new patchwork model against older methods that tried to describe the entire year of wave data with just one single formula. They found that while a single formula might work well for a season that is dominated by just one type of wave, it fails to capture the full picture when the ocean shifts between calm, moderate, and extreme states. The new spliced model proved to be a better fit for the data, particularly because it could explicitly separate the dangerous tail events from the routine daily waves. This separation is not just a statistical trick; it translates directly into better decision-making. By knowing exactly which season brings the highest risk of extreme waves and which season offers the longest windows of calm, operators can plan their logistics, schedule their maintenance, and design their pipelines with greater confidence. The study concludes that for the Abadi-1 Masela area, understanding the ocean requires looking at it in parts rather than as a whole, ensuring that the safety of the people and equipment working on the water is based on a realistic view of the sea's true nature.

Technical Summary: A Three-Component Spliced Lognormal-Weibull-K Distribution Framework for Seasonal Modeling of Significant Wave Height in the Abadi-1 Masela Offshore Area

Problem Statement
Significant wave height ( $H_s$ ) is a critical oceanographic parameter for offshore operations, including drilling, subsea pipeline risk assessment, and weather window definition. However, traditional modeling approaches often rely on single probability distributions (e.g., Lognormal, Weibull, Rayleigh) to characterize the entire range of wave data. The authors argue that single-distribution models fail to adequately capture the high variability inherent in wave-height statistics across three distinct regimes: low-wave, moderate-wave, and upper-tail (extreme) events. This limitation is particularly relevant in the Abadi-1 Masela offshore area, where monsoon-driven seasonality creates distinct wave climates that a single annual average or distribution cannot accurately represent. Consequently, there is a need for a regime-based probabilistic model that can flexibly adapt to seasonal variations and distinct wave conditions to support operational safety and risk assessment.

Methodology
The study proposes a Three-Component Spliced Lognormal-Weibull-K Distribution framework tailored to the seasonal characteristics of the Abadi-1 Masela offshore area. The methodology involves the following steps:

Data Source and Preprocessing: The analysis utilizes 3-hourly significant wave height ( $H_s$ ) data (variable VHM0) from the Copernicus Marine dataset, covering the period from 2022 to 2026. Data were screened for validity and positivity ( $H_s > 0$ ).
Seasonal Classification: The dataset was segmented into four seasonal periods based on the monsoon cycle: West Monsoon (December–February), Transition I (March–May), East Monsoon (June–August), and Transition II (September–November).
Wave Regime Definition: The $H_s$ $H_{s}$ data were classified into three specific regimes to facilitate splicing:
- Low-wave regime: $H_s \leq 1.5$ m
- Moderate-wave regime: $1.5 < H_s \leq 2.5$ m
- Upper-tail regime: $H_s > 2.5$ m
Distribution Assignment:
- The Lognormal distribution was assigned to the low-wave regime.
- The Weibull distribution was assigned to the moderate-wave regime.
- The K-distribution was assigned to the upper-tail regime.
Model Construction: A spliced probability density function (PDF) was constructed by concatenating these three distributions. The weight of each component ( $w_1, w_2, w_3$ ) was calculated based on the proportion of observations falling within each regime for a given season. Parameters for each distribution were estimated using Maximum Likelihood Estimation (MLE).
Model Evaluation: The performance of the proposed spliced model was compared against single-distribution models (Lognormal, Weibull, and K-distribution) using four goodness-of-fit metrics: Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), Root Mean Square Error (RMSE), and the Kolmogorov–Smirnov (KS) statistic.

Key Results
The analysis revealed significant seasonal variability in the wave climate of the Abadi-1 Masela area:

East Monsoon: Exhibited the highest mean $H_s$ (1.539 m) and a nearly balanced distribution between low-wave (48.37%) and moderate-wave (49.32%) regimes, indicating the most energetic sea conditions.
Transition II: Represented the calmest conditions with the lowest mean $H_s$ (0.929 m). It was dominated by the low-wave regime (89.48%) and contained no observed upper-tail events ( $H_s > 2.5$ m).
West Monsoon: While dominated by low waves (75.31%), this season recorded the highest maximum $H_s$ (3.870 m) and the highest proportion of upper-tail events (4.12%), making it the period of greatest extreme-wave risk.
Model Performance: Goodness-of-fit results varied by season. In seasons dominated by a single regime (e.g., West Monsoon dominated by low waves), single distributions (Lognormal) sometimes yielded lower AIC/BIC values. However, the spliced model consistently provided a more physically interpretable representation of the data structure. Notably, for Transition II, the spliced model was the best fit across all metrics (AIC, BIC, RMSE, KS), demonstrating its adaptability by effectively excluding the K-distribution component when upper-tail data were absent.

Significance and Contributions
The paper claims that the primary contribution of this work is the development of a regime-based spliced probabilistic model that offers greater flexibility and interpretability than customary single-distribution models. The significance of the study lies in:

Operational Relevance: The model explicitly links statistical distributions to operational states: low-wave conditions correspond to normal operations, moderate-wave conditions to restricted operations, and upper-tail events to high-risk safety scenarios. This distinction is crucial for offshore risk assessment, weather window planning, and pipeline design.
Seasonal Specificity: By modeling each season separately, the framework captures the distinct probabilistic structures of the Masela Block's wave climate, which a single annual model would obscure.
Adaptability: The framework is shown to be adaptive; it does not force a tail distribution (K-distribution) onto seasons where extreme events are not empirically present (e.g., Transition II), thereby avoiding the imposition of unsupported statistical assumptions.
Risk Characterization: The study provides a robust basis for characterizing seasonal offshore wave risks, specifically highlighting that the West Monsoon, despite not having the highest mean wave height, poses the highest risk due to the frequency of upper-tail events.

The authors conclude that while single distributions may suffice when a single regime dominates, the three-component spliced framework is superior for comprehensive offshore risk analysis as it accurately reflects the multi-regime nature of significant wave height in the Abadi-1 Masela offshore area.

Technical Summary: A Three-Component Spliced Lognormal-Weibull-K Distribution Framework for Seasonal Modeling of Significant Wave Height in the Abadi-1 Masela Offshore Area

More like this