coherence map

A coherence map in 3D seismic interpretation is a two-dimensional image extracted from a three-dimensional coherence volume that displays lateral waveform similarity between adjacent seismic traces on a time slice, depth slice, horizon-extracted surface, or interval-averaged window, producing a grayscale or color image in which values approaching 1.0 indicate continuous laterally coherent reflections from intact formation and values approaching 0.0 indicate abrupt waveform changes at faults, fractures, channel boundaries, karst collapse features, and stratigraphic discontinuities; the three underlying coherence algorithms (cross-correlation C1, semblance C2, and eigenstructure C3) all produce volumetric coherence values that are displayed as coherence maps by selecting any horizontal, vertical, or dip-following extraction plane through the 3D seismic cube, with the choice of extraction plane determining which geological features are imaged most clearly. In the Western Canada Sedimentary Basin, coherence maps extracted from 3D seismic surveys over Cretaceous Mannville, Viking, and Cardium formations, Devonian Leduc, Nisku, and Swan Hills carbonate reefs, and Triassic-Jurassic Montney and Doig siltstone plays are standard deliverables from seismic interpretation programs, enabling WCSB exploration and development teams to map fault systems that control trap integrity in Devonian reef plays, delineate Cretaceous incised valley channels that guide SAGD and CSS horizontal well steering in Athabasca and Cold Lake heavy oil pools (where channels 200 to 800 m wide must be resolved at 25 m trace spacing), and characterize natural fracture corridors in Montney and Duvernay unconventional plays where fracture orientation and density govern hydraulic fracture propagation across laterals of 1,500 to 3,000 m; WCSB 3D surveys used for coherence mapping range from 50 km2 to 5,000 km2, with C2 semblance as the Alberta Plains standard (gentle dips of 1 to 5 degrees) and C3 eigenstructure reserved for WCSB Foothills programs where steep dips exceed C2 algorithm performance.

• Time-slice versus horizon-extracted coherence maps in WCSB interpretation: The choice of extraction plane determines which geological features are imaged most effectively on a coherence map, and WCSB interpreters routinely generate both types for each target horizon. Time-slice coherence maps at a constant two-way travel time cross all formations at the same depth level and are most effective for imaging near-vertical faults, fracture zones, and channel systems that intersect the slice at high angles, appearing as continuous linear or curvilinear lows; WCSB Viking Formation tight oil channel systems at Pembina, Garrington, and Swan Hills fields are commonly mapped on time slices at 400 to 900 ms because the channels have near-vertical margins with respect to 25 m trace spacing and appear as sharply defined ribbon anomalies 200 to 800 m wide with coherence values of 0.55 to 0.72 versus 0.87 to 0.93 in inter-channel shales. Horizon-extracted coherence maps follow the interpreted seismic reflection from a specific formation top, extracting coherence along reflector dip rather than on a flat time slice, and are most effective for imaging features parallel to the stratigraphic surface including reef margins, channel bases, and diagenetic fronts; in WCSB Devonian Leduc reef plays at the Rimbey-Meadowbrook and Grosmont trends, horizon-extracted coherence maps at the reef top show the circular to elongate outline of the reef complex as a high-coherence interior (0.88 to 0.95) surrounded by a coherence-low margin (0.45 to 0.65) at the reef-to-off-reef transition, which controls trap geometry and water influx pathways in mature Devonian waterflood fields.
• Coherence map display conventions, color rendering, and interpretation pitfalls in WCSB programs: Coherence map display in WCSB interpretation platforms (Petrel, Kingdom, OpendTect) renders low coherence in black or dark blue (discontinuities, faults, channels) and high coherence in white or yellow (continuous reflectors, intact formation), with the color scale calibrated to the specific coherence algorithm and data quality of the survey; C2 semblance maps for WCSB Alberta Plains surveys typically show background coherence values of 0.85 to 0.95 in undisturbed formations, with fault anomalies dropping to 0.40 to 0.65 and channel boundaries at 0.55 to 0.75. The most common interpretation pitfall in WCSB coherence mapping is conflating acquisition footprint artifacts with geological discontinuities: receiver-line and source-line spacing in WCSB 3D surveys (200 to 300 m in both directions) can produce coherence lows aligned with the survey geometry that mimic fault lineaments, particularly in shallow intervals (above 500 ms) where data fold is lower and bin-to-bin coherence variation is larger; distinguishing acquisition footprint from genuine faults requires confirming that the anomaly orientation differs from the survey line direction by more than 15 degrees and that the anomaly persists across at least three to five consecutive time slices rather than appearing only on the slice that maximizes the fold variation pattern.
• Coherence map integration with amplitude and curvature in WCSB multi-attribute interpretation: WCSB interpretation teams routinely co-render coherence maps with amplitude and curvature attributes in blended RGB or opacity-weighted displays to distinguish fault-controlled discontinuities from stratigraphic amplitude variations from diagenetic boundaries. In WCSB Cardium tight oil plays at Pembina, Garrington, and Buck Lake fields, co-rendered coherence and amplitude maps on the Cardium horizon distinguish sand-filled channels (high amplitude, low coherence at margins) from shale-plugged channels (low amplitude, low coherence at margins) from inter-channel shales (low amplitude, high coherence), providing the lithological context to prioritize horizontal well targets within the coherence-imaged channel system; at Pembina Cardium, this multi-attribute approach reduced dry hole rates from approximately 18 percent to less than 7 percent compared to amplitude-only mapping by eliminating shale-plugged channel targets. Curvature attributes (most positive, most negative, shape index) highlight reflector bending from faulting and compaction and are overlaid on coherence maps in WCSB Montney and Duvernay interpretation to identify fracture corridors where both a coherence low (impedance contrast from the fracture zone) and a curvature anomaly (mechanical bending at the fracture swarm) coincide, confirming the fracture as a structural feature rather than a noise artifact, with WCSB Duvernay operators reporting 25 to 40 percent improvement in completion performance in wells targeting coherence-and-curvature coincident fracture corridors compared to wells targeting coherence-only anomalies.
• Coherence map quality control and noise suppression in WCSB 3D interpretation workflows: The quality of a coherence map is sensitive to seismic signal-to-noise ratio and pre-processing steps applied before coherence computation; WCSB 3D interpretation workflows routinely apply dip-steered coherence filtering (structure-oriented filtering) to the seismic volume before coherence extraction to suppress random noise that otherwise produces false coherence anomalies at the trace-spacing scale (12.5 to 25 m) mimicking genuine geological discontinuities at the 50 to 200 m scale of Mannville channels or Montney fracture corridors. Coherence map quality control in WCSB production interpretation programs uses three metrics: minimum fault detection length (the shortest coherence anomaly consistently mapped by two independent interpreters working blind, typically 200 to 500 m at WCSB Alberta Plains signal-to-noise ratios of 3 to 8 dB); coherence threshold for edge definition (the value below which a pixel is classified as a discontinuity edge, calibrated by comparing coherence maps to production data from wells drilled into channel margins, typically 0.70 to 0.80 on C2 semblance maps); and noise floor estimation from coherence maps over known-continuous formations (flat marine shales in WCSB Cretaceous sections provide a controlled measurement of background coherence variation from noise alone, typically 0.03 to 0.08 coherence units RMS).
• Coherence map use in WCSB 4D time-lapse monitoring and SAGD steam chamber imaging: In WCSB Athabasca oil sands SAGD operations, coherence maps extracted from time-lapse 3D seismic surveys (acquired annually or biennially over producing SAGD pads at Cenovus Foster Creek, MEG Energy Christina Lake, and Canadian Natural Resources Jackfish) image the expanding steam chamber as a zone of reduced coherence relative to cold bitumen surroundings, because steam-heated bitumen has a markedly different seismic impedance from unheated cold bitumen (P-wave velocity drops from 3,100 to 3,500 m/s in cold bitumen to 1,800 to 2,400 m/s in steam-saturated sand at 250 to 270 degrees Celsius), producing a coherence anomaly at the steam-cold bitumen boundary that can be tracked between survey vintages to measure lateral steam chamber growth rate. WCSB SAGD coherence maps from 4D surveys are used operationally to identify steam breakthrough risks where the coherence anomaly extends vertically to within 20 to 40 m of the overburden caprock or laterally toward adjacent well pairs at 50 to 100 m inter-well spacing, triggering injection rate adjustments before breakthrough; at Foster Creek, 4D coherence monitoring has extended average well pair production plateau periods by 1 to 2 years compared to operations without seismic monitoring by enabling earlier detection of steam chamber merging between adjacent pairs, validated against production rates and temperature observation well data at 12 reference well pairs across the Foster Creek expansion program.

Coherence Map Resolving WCSB Viking Channels at Garrington Field

A WCSB operator at Garrington field in central Alberta used a horizon-extracted C2 semblance coherence map on the Viking Formation to plan a 5-well horizontal program targeting braided fluvial channels at 850 m depth. The coherence map, extracted on a 4 ms window straddling the Viking Formation top, resolved channel margins at 200 m width with coherence values of 0.64 to 0.71 compared to 0.89 in inter-channel shales. Two channels previously mapped as a single wide anomaly on amplitude data were resolved as separate strands with a 120 m inter-channel shale plug by the coherence map, redirecting one planned well trajectory by 400 m to remain within productive sand. Post-drilling correlation at 5 well control points confirmed coherence-imaged channel boundaries within 50 m in 4 of 5 cases. Average 6-month production from the 5 coherence-guided wells was 38,000 barrels of oil, compared to 24,000 barrels for the preceding 4 amplitude-guided wells in the same field, a 58 percent improvement attributed to more precise lateral placement within the channel axis using the coherence map.

Related Terms

Coherence is the parent attribute; a coherence map is the 2D display produced by extracting the coherence volume on a specified time slice or horizon surface; the C1, C2, and C3 algorithms and analysis parameters (window length, trace aperture) that generate the coherence volume directly control coherence map image quality. Coherence filtering (structure-oriented or dip-steered filtering) is applied to the seismic volume before coherence map extraction in WCSB workflows to suppress random noise that generates false discontinuity anomalies at trace spacing scale, improving geological signal on Mannville and Viking coherence maps. Coherence vector map adds directional information to the scalar coherence map by computing the spatial gradient of coherence anomalies, displaying fault strike and dip as vector arrows for quantitative fault orientation mapping across WCSB 3D survey areas. Seismic attribute is the broader category encompassing coherence maps, curvature, spectral decomposition, and AVO outputs; coherence is the most widely used discontinuity attribute in WCSB 3D interpretation programs. Time-lapse seismic (4D) uses coherence maps from repeated 3D surveys to track WCSB SAGD steam chamber expansion; the velocity contrast at the steam-cold bitumen boundary produces a coherence anomaly monitored annually at Foster Creek, Christina Lake, and Jackfish.