Remembering to turn on the tap

I've been working on an update to the LIFE pipeline for a few months now, which has caused me to refactor some of the pipeline, including moving some stages that were in Python code to just calling GDAL command line tools instead. What follows is an error in translation in that process, but I thought I'd document it as I think it's not a flag that I see used often, but can have a real impact on your results despite not seeming to do much. When I sent the most recent version of the LIFE maps from my revised pipeline out to the group for review one of the group PhD students, Emilio Luz-Ricca, spotted an interesting oddity, that whilst most of the oceans are converted to NODATA, there was this odd ring around all the land masses containing zero values. Emilo also observed that it happened to coincide with the transition between Marine (Neretic) and Deep Ocean Floor in the Jung habitat map we use: As I mentioned, all the values in this mystery buffer were zero, so in theory have no…