Passive margin clastics EAGE 2026 preliminary Open-source dataset

Penobscot: 2D-to-3D horizon reconstruction with calibrated uncertainty.

A rare combination — sparse 2D lines and a real 3D survey acquired over the same area — used as the one-time opportunity to validate a reconstruction method against true ground truth.

Location
Offshore Nova Scotia
Dataset
11 × 2D lines + 3D volume
Method
RBF prior + GP posterior
Validation
4 horizons, plan-view + section

Why Penobscot

The Penobscot dataset contains both a 3D seismic volume and a set of 2D seismic lines acquired over the same area with comparable acquisition. Real 2D lines, real wavelet, real noise — and a real 3D ground truth to validate against.

That combination is rare. It lets us answer one specific question: given sparse 2D picks of a horizon, can we reconstruct the 3D surface, and how well do our uncertainty estimates match the actual error?

The workflow

The reconstruction starts from 2D horizon picks delivered by the calibration project. Each surface is reconstructed two ways — by RBF (the classical smooth prior) and by GP regression (which produces both a mean estimate and a posterior covariance). This QC notebook then compares both reconstructions against the held-out 3D ground truth.

Penobscot reconstruction workflow diagram
Figure 1 · Workflow
Reconstruction data flow. Customer data flows through imports into 2D and 3D horizon stores; 2D horizons are reconstructed per-surface via RBF (03*) and GP (04*); the 3D ground truth is held out for QC.

Section-view comparison

The headline visual: ground truth versus both reconstruction methods, on two held-out crosslines and four reconstructed horizons. The GP P10–P90 envelope is visible as light blue shading around the GP mean — narrow at line locations, wider between lines.

Section view of reconstructed horizons at XL 1165 and XL 1326
Figure 2 · Section view at XL 1165 & XL 1326
Black: OD 3D ground truth. Orange dashed: RBF tied. Blue: GP mean with P10–P90 shading. Surfaces, shallow to deep: Seabed, Logan Canyon, Base-O Marker, Baccaro.

Plan-view comparison

Plan-view two-way-time heat maps for all four horizons, hill-shaded for structural relief. Color scale is fixed per horizon across all three panels so the visual comparison is fair.

Plan-view TWT heat maps for four reconstructed horizons
Figure 3 · Plan view, all four horizons
Left: 3D ground truth. Middle: RBF tied. Right: GP mean. Dashed white lines mark the reference crosslines (XL 1165, XL 1326).

Error analysis

Per-pixel plan-view error maps (reconstruction − ground truth) for both methods. The summary statistics tell the story: GP and RBF perform comparably on three of the four horizons. Baccaro is an outlier on both methods, with a ~30 ms bias that traces back to the well-documented processing mismatch between the 2D and 3D acquisitions.

Horizon RBF bias RBF std GP bias GP std GP wins
Seabed −0.7 ms 13.1 ms +0.4 ms 11.1 ms 58.7%
Logan Canyon +4.3 ms 10.9 ms +3.1 ms 13.6 ms 42.8%
Base-O Marker −6.9 ms 9.1 ms −6.8 ms 9.5 ms 48.8%
Baccaro +29.1 ms 15.1 ms +30.8 ms 15.0 ms 43.9%
Per-horizon error maps for RBF and GP reconstructions
Figure 4 · Error maps, RBF and GP
Plan-view (reconstruction − ground truth) per horizon. Bias and standard deviation annotated in each panel title.
Per-pixel GP-vs-RBF improvement map
Figure 5 · GP − RBF improvement
Positive (blue) = GP beats RBF. GP wins clearly on Seabed (58.7%); the other three horizons are roughly a coin flip.

Calibration

A point estimate without a calibrated uncertainty is half a result. We test the GP's reported uncertainty by measuring what fraction of ground-truth points fall inside the GP mean ± 1σ band. For a well-calibrated posterior, that fraction should land near ~68%.

Horizon GT inside GP ± 1σ Target Reading
Seabed 55.5% ~68% Slightly under-covered
Logan Canyon 62.7% ~68% Close to calibrated
Base-O Marker 39.7% ~68% Under-covered — σ too tight
Baccaro 5.8% ~68% Badly under-covered — see limits below
GP posterior coverage maps per horizon
Figure 6 · GP posterior coverage
Plan-view fraction of GT inside (GP mean ± σ) per horizon. Logan Canyon is the best-calibrated of the four; Baccaro is badly under-covered, consistent with the systematic 2D-vs-3D time-domain mismatch documented in the pre-flight QC.

Honest limits

What this preliminary result does and does not show

On Seabed and Logan Canyon the GP reconstruction is competitive with — and on Seabed clearly better than — the RBF baseline, with sigma calibration close to target. That is the expected signature of a working method on well-behaved horizons.

On Base-O Marker and Baccaro the GP posterior is too narrow: the reported uncertainty understates the actual error. The Baccaro miscalibration traces to a known systematic mismatch between the Penobscot 2D and 3D acquisitions (independent processing, ~12 ms median time shift, 4.7× amplitude scaling, different wavelets). Until that is corrected per-line, no reconstruction method should be expected to land its uncertainty bands honestly on Baccaro.

Penobscot answers one question: does the method work in principle when ground truth exists? The early answer is yes — on the horizons where the input domain is consistent. The same framework, deployed without ground truth, is addressed in the commercial deployment work.

Acknowledgments

Penobscot is a publicly released dataset made available through the Nova Scotia Department of Natural Resources. The 2D and 3D seismic data, well logs, and interpretations used in this case study are open-source and freely redistributable for research and demonstration.