Testing a Deep-Source Branching Hypothesis Across Petroleum, Coal, Hydrothermal, and Evaporite–Brine Systems

Petroleum accumulations, coal-bearing successions, hydrothermal fluids, and evaporite–brine systems are usually investigated in separate disciplinary literatures. This preprint develops a falsifiable cross-system framework to test whether these domains can participate in a shared deep-connected architecture without assuming a common ultimate material source. A deep-to-shallow model is tested as the active hypothesis, while empirical direction remains evidence-gated and may be supported, rejected, reversed, or left unresolved. The framework separates measurement from author interpretation and from the present synthesis and applies the sequence fingerprint → connection → chronology → direction. Literature evidence from the Michigan, Permian, and Sichuan basins identifies a recurrent candidate petroleum–hydrothermal–evaporite/brine architecture, although the strength of the individual links is uneven and overall cross-basin closure remains PARTIAL. A focused test of the Longtan coal-bearing domain yields B8 = STRONG PARTIAL PASS. Coal is classified as a candidate connected transformed branch, not as a proven common-source or deep-source product. Same-source resistant-tracer closure and exact same-conduit closure remain OPEN. The synthesis supports a branching, filtering, reaction-coupled and multi-pulse architecture compatible with deep contribution, but it does not demonstrate a single common ultimate source for petroleum, coal, hydrothermal fluids and evaporite–brine systems.