Many deep-marine sedimentological studies have been published, spanning modern, subsurface and outcropping systems. The ability to integrate this growing volume of data would allow new insight into deep-marine system organisation to be developed. Yet to date, comparative analysis of deep-marine studies has been hindered by the wide variety in methods of data collection, scales of observation, resolution, classification approaches and adopted terminology. To address this variety, a relational database approach has been designed and implemented. Known as the Deep-Marine Architecture Knowledge Store (DMAKS), it offers a means to standardise data acquisition, characterising boundary conditions, together with architectural and facies properties, in a spatial, temporal and hierarchical framework.
Three key work elements are presented. The first is a critical review of deep-marine hierarchical classifications, covering the principles that commonly underpin them, their history of development and the degree to which they can be reconciled. The second comprises a series of DMAKS applications, including (i) channel geometry characterisation, (ii) review of hierarchical organisation of channelised and terminal deposits, (iii) assessment of temporal trends in terminal lobe deposition, (iv) development of scaling relationships between adjacent channel and levee architectural elements, (v) quantification of the likely occurrence of elements of different types as a function of the lateral distance away from a known point, (vi) evaluation of proportions and transition statistics of facies in elements and beds, (vii) characterisation of variability in net-to-gross ratios among element types. Thirdly, a comparative study assesses the influence of external controls on the development of lobate terminal deposits at multiple depositional scales.
The DMAKS database approach is shown to represent an advance on previous deep-marine databasing efforts due to the breadth of its scope and its capacity to characterise deep-marine organisational styles. DMAKS has the capability to i) test facies and architectural models against a wider data pool, ii) tailor quantitative outputs to suit a specific environment through filtering on combinations of stored parameters and iii) establish predictive models and statistically-supported synthetic analogues through comparison with analogous datasets; these all may reduce geological uncertainty in reservoir models. The value of such database outputs arguably augments the value of existing deep-marine sedimentary data; the standardised data contained in DMAKS has the potential to yield deeper understanding than that which can be derived from individual studies alone.
