As sedimentologists, it is important not "To see the world in a grain of sand", in the words of William Blake. Often it is best to temper our interpretation to the lowest common denominator, rather than making sweeping statements about a depositional setting based on more scarce data which is deemed to be of higher quality or accuracy. A depositional event which is recorded in one locality only may have numerous potential drivers, which may not be applicable over a larger area.

To add another well-known quote "The absence of evidence is not evidence of absence". Using "missing/absent" data as a control for depositional character over a regional scale can cause misrepresentation. Removing assumptions like this, missing data means a unit is missing for example, is a must.

A key concept is time averaging. Even when your data has excellent age control (usually with biostratigraphy), the display is most commonly at stage or substage level resolution. That still means a couple of millions of years in most cases. One has to consider the most representative paleoenvironment at that location in the stratigraphical interval of interest to be displayed. The vertical and horizontal scale of averaging will depend on the resolution of your data.

A depositional environment map spatially captures an interpretation of depositional environment for a defined interval of geologic time. The interpretation should be based on data that are as close to age equivalent as possible. The whole point is to limit temporal variability in a defined way and then capture spatial variation in depositional environment on a map. Problems ensue when the geologic age or age equivalence of the data is incorrect, or if the data is incorrectly located.

The scale and variability of depositional environments based on physical controls alone requires an understanding of scale. The environment of deposition must be mapped based on data control points where core or other measured data is available. Control on the temporal variations down section should then control the mapped zones to ensure the depositional variations are scalable and comparable within the same ages. An example being the present day earth surface. There are countless examples of lateral variations in depositional events at a scale that could be missed if assumptions are applied with no data control. These analogue reviews can keep things on track. Check your scale. Check you stratigraphic interpretations.

Indeed with Mr. Gil Machado, and also another key factor is what is represented on the map of the paleo environment, it is only an instant photo in a given time of the paleo environments represented.

Sometimes the simplest is to assume the lack of knowledge in some areas and/or intervals and use question marks or other notation. It is also important to differentiate between actual data and interpretation - different shades in a map are a good way to do it.

Observations should be solid data points, but as soon as a worker goes into the field or to a core lab with a model in mind, seeking to "prove up the model", science hits a wall at 100 mph. They will see what they want to see and may miss key observations that would disprove the model. This is an issue within numerous widely referenced papers by highly regarded workers. Multiple hypotheses are required and an open mind and keen eye toward all possible observations and interpretations is important. Finding even one observation that casts doubt on your preferred interpretation is the observation that must be addressed rather than ignored.

It is especially important in energy transition workflows to temper the way we present gross depositional environment maps for stakeholder engagement. GDE maps are inputs into Common Risk Segment maps used extensively in oil and gas settings, and increasingly in exploration for low-carbon initiatives such as carbon storage or nuclear waste disposal. Variations in sediment type or 'quality' may lead to a Common Risk Segment map appearing as red instead of green. Red is an alarming colour and can often lead people to assume an opportunity or asset is immediately unviable. In this regard, effective geocommunication in the way we construct depositional maps for presentation to non-specialists is crucial.

One can oversimplify as is often the case with presentations to non-geologists. This may lead to expensive investments with catastrophic outcomes. An example might be injecting CO2 into a deltaic reservoir with a broad areal extent - this sounds great. Wave-dominated deltas can be huge, fairly uniform sheets, but river-dominated deltas have tremendous heterogeneity. The term "Delta" isn't enough and at least one highly-publicized CCS project has failed for this reason.

The old adage that "They who have seen the most rocks win" is accurate, but I would add that "They who have seen the most rocks, thought carefully about them and revisited the exposure multiple times with other colleagues to test their interpretation might win".

Paleoenvironmental interpretation emerges from analogies with other study objects and datasets. Therefore, the background/experience of the interpreter is crucial. The more rocks you see (preferentially in the field) the better.