The old and new of fluid indicators; an overview and application; discussion

The matter I want to raise in this note is about the "background trend" mentioned in Farfour and Russell (2024). It has been shown that this is an artifact (de Bruin, 2019, 2020a). The authors refer to a number of publications, including several by accomplished geophysicists, who all got this "trend" wrong in the 1980s and 1990s. The list includes names like Castagna, Goodway, Hilterman, Foster, Connolly, and Russell. Regardless of all the rewards those papers have accumulated, if it's wrong then it's wrong. The success of amplitude variation with offset (AVO) fluid indicators has nothing to do with the "well-defined background trend" referred to in Farfour and Russell (2024): "Seismically, the success of the AVO fluid indicators stems from the fact that the extracted AVO intercept and gradient (or RP and RS), in the absence of hydrocarbon, often exhibit a well-defined background trend (Castagna and Smith, 1994)." The authors omit reference to two other papers, one by Cambois (1998) and one by Hendrickson (1999), who for the first time (at least in the open literature) suggested that this trend might not be caused by lithology. Those references should have been included as they had a major impact on the way people looked at AVO and fluid indicators. Also not referenced are de Bruin (2019, 2020), which explain how on closer examination the "background trend" turns out to be a mathematical artifact. Some will perhaps say that it does not matter as long as it helps to detect hydrocarbons, but that is hardly a scientifically justifiable approach. The old and incorrect theories are presented as cause and effect. Now the cause has been shown not to exist, but it seems that we do not want to let go of the effect. The effect can be found (because it is in the data) in other ways. To have this artificial trend, sometimes called fluid line, as an intermediate step is completely arbitrary. Others have said that there may be an artificial contribution to the trend, but real rock properties contribute to it as well. This does not make sense either. The artificial trend dominates what you see in intercept-gradient crossplots. If there would be a contribution from rock properties, then it is not a good idea to contaminate it with such a strong artificial trend. This artifact has become an important but unnecessary ingredient of many AVO theories, and I believe it is time to abandon it. The way in which amplitudes vary with offset is very important, and the fact that they do so is very useful. There are ways to benefit from this without the intermediate step of a "background trend". A way to do so is described in de Bruin (2020b). Although this matter is well explained in referenced papers, I include a brief summary here.