Hydrotowns 3: Habitability and niches
The starting point for this side-track was coming across an article about what the authors term the "human climate niche". It's open access and available here. I'm certainly not an expert in this field so I can't comment on whether the work that they show is missing some critical features but I will try to highlight what caught my attention.
The main figure shows a variety of parameters as a function of mean annual precipitation (MAP) and mean annual temperature (MAT). These two factors seem to have done the heavy lifting in the past when people have looked into things like habitability, biomes and so on. For example, the "Whittaker diagram" is a commonly referenced biome classification which uses these parameters. I suppose its main use case is the fact that it is just a simple mapping. More sophisticated classifications like Köppen also take into account the seasonality of rainfall and temperature using some fairly convoluted rules.
As an aside, the impact of seasonality is one which can be overlooked. For example, places like Berlin and Toulouse actually have similar MAP values to Cape Town. In general it seems to be accepted that scrub/macchia vegetation like fynbos develops in areas that are bioclimatically suitable for temperate forests, but where the combination of hot and dry summers means that these areas are shaped as much by fire as by the overall precipitation or the average temperature.
In fact, in areas where fire has been excluded by human means (e.g. Orange kloof at Hout bay), it has been observed that forest patches have started to expand out of their so-called fire refugia into the fynbos. As an aside of an aside this also probably goes some way to explaining why other natural areas are easily invaded by pines and wattles. More on fire and the land ethic probably in the future.
In any case, going back to the original paper, it aimed to explore the existence of a human "niche" with respect to MAT and MAP - i.e. whether human population density is distributed in a particular way with respect to these parameters. As their data shows, this appears to be true, with the peak lying somewhere around a MAT of 13C and a MAP of 1000mm. They also demonstrate the existence of an additional peak at a higher temperature and precipitation level which apparently corresponds to the Indian monsoon region, although I imagine it also includes other regions with similar climates. The figure below shows the section of this distribution along the MAT axis.
They seem to be more interested in the temperature relationship, as they then show how this could affect the distribution of people under various climate change scenarios (the RCPs from the IPCC's fifth assesment report, it would be interesting to see how they compare to the new SSP trajectories). It's not necessarily a projection per se but merely shows how population density would need to change to maintain the same niche distribution as the climate changes. In the supplementary data, however, they also show how this would change when taking into account both the temperature and precipitation contributions to the niche. To me this seems a bit more logical (thinking back to the climate classification story above) - in the paper they mention that it mainly affects desert regions but to me it seems like it has a more striking effect. I placed both figures next to one another, you can judge for yourself. They are apparently on the same scale.
To me there are some differences, particularly in the intensity. In any event, I think it's an interesting analysis and one I would like to take forward in some way. To me there are two questions that arise: firstly, what other factors might be important? Secondly, can we come up with a single index which can take into account the factors which contribute to this niche? I came across one such index in the past which was designed to show areas where the climate is mild enough for temperate rainforests (with ferns) to occur. I get the feeling this kind of index could be used to create something a bit more sophisticated, particularly if we limit the region we are looking at. I hope to look into this in the future and maybe come up with some similar figures at a finer and more local scale.
DW