Aquamaps is this cool website where the species of marine organisms are given automatically generated distribution maps based on the environmental factors of their occurrence locations, such as salinity, temperature and depth. These can then be further refined by expert review to be more correct. It works for all taxa, from mollusks and arthropods to fish and whales to kelp.
Which begs the question, why is there is no equivalent Terramaps for terrestrial organisms? different set of environment factors would have to be used (vegetation cover, soil pH, elevation, humidity, etc), but I don’t see why the same process could not be applied. You may say lack of demand is a huge issue, after all most terrestrial vertebrates already have their ranges mapped. To which I say, perhaps these could produce more accurate maps. And then there is the fact there is a whole lot of invertebrates, fungi, and plants who’s ranges are not mapped.
It sounds like this is a modelling exercise producing maps of where the environment may be suitable for particular species rather than distribution maps. The latter can only be generated by records.
But they wouldn’t be distribution maps. They could be useful for guiding you in where to look for under-recorded species.
Niche mapping using the process you describe has existed for decades, and has certainly been applied to both terrestrial and aquatic organisms very many times. It is not more widely known because it inevitably oversimplifies what factors determine the range of any species. Models often don’t include barriers to dispersal, relationships with other organisms, interactions with human infrastructure, dependence on micro-climates, etc. There are also problems with spatial resolution, as when this is done for many species it tends to have a fairly coarse grid, but there is lots of environmental variation within those big grid squares.
I’m nor sure what point you are making. Obviously the occurrence records have to come into the process in order for the modelling to work out where else the environment appears to be suitable for the species to occur. But I’m not sure this makes the distribution maps more accurate, which is what you suggested.
I suppose the modelling could highlight outlying records, outside what appears to be the environmental tolerance of the species, and if these prove to be misidentifications and are removed from the species dataset, that would make the distribution maps more accurate. Is that what you had in mind? It would require re-examination of specimens because it would be difficult to judge whether an outlier was showing the limit of environmental tolerance or was a rogue record outside the tolerance limits.
As dlevitis says, however good the environmental modelling, there will always be other factors that could be taken into account and they may explain why a species is absent from apparently suitable habitat. I don’t think it would ever be safe to say “This species hasn’t been found at X but the environment is suitable so we should include X in the species’ distribution map”
It would require re-examination of specimens because it would be difficult to judge whether an outlier was showing the limit of environmental tolerance or was a rogue record outside the tolerance limits.
That would be for expert revision to decide on a case by case basis.
As dlevitis says, however good the environmental modelling, there will always be other factors that could be taken into account and they may explain why a species is absent from apparently suitable habitat. I don’t think it would ever be safe to say “This species hasn’t been found at X but the environment is suitable so we should include X in the species’ distribution map”
Well, no method of range mapping is truly perfect. But its a great starting place