Current warming and likely future impacts - Climate Change: Biological and Human Aspects - page 297

the Hadley model suggested that part of the Sierra Nevada will experience drought.
Both models do indicate that Montana, Wyoming and North and South Dakota may
retain less soil moisture, as would be expected from looking at past climates when it
was a little warmer than today.
These temperature and rainfall factors suggest increases in vegetation with climate
change across much of the USA, with the exception of the Mississippi floodplain
and Florida. However, the Hadley model suggested a marked decline in vegetation in
Arizona. This was not as big a difference between the two models as it seems. If an area
is already vegetation-poor, in this case due to aridity, then only a small proportional
change in an already small biomass per unit area equates to a big difference in the
ecosystem. So much depends on which parameter is represented more in the output.
At the moment one of the larger features in an ecosystem map (technically, a
biome map) of the USA is the eastern states, with their temperate deciduous and
mixed forest, and the grasslands of Oklahoma, Kansas, Nebraska and South Dakota,
with a border between these two ecosystems running through eastern Texas, and up to
the Iowa/Illinois boundary. This boundary line lies north by north-east. However, at
the end of the 21st century both models suggest that this will become north-south, still
beginning in mid-Texas but instead ending close to the Minnesota and North Dakota
border (see Figure 6.4). These last two states, together with Iowa and Missouri, will
see considerable ecosystem change.
These are the broad changes across large areas that both models predict: there
will also be climate change elsewhere. However, they also predict much change (but
differently) on smaller scales. The Rocky Mountains, already a highly fragmented
natural landscape in terms of ecosystem type, are predicted to change. Nonetheless,
it is possible that the small lateral distances of such change might be accommodated
through vertical migration, but the picture is not that clear for definitive predictions,
let alone the national determination of ecosystem-management policy at the local
level.
The agreement, or lack of, between the two models used in the first USGCRP
report from 2001 is aptly illustrated by how they each predict the summer climate of
Illinois will change by 2090 (see Figure 6.4). Both models predict that the climate
approximately along the latitude 35 ◦ N will shift 5 ◦ further north by the end of the 20th
century. However, they disagree as to whether Illinois's climate will become like North
and South Carolina's or will be more like Oklahoma's. The main difference between
the two is the amount of summer rainfall: Oklahoma gets less.
With regards to ecosystems, the 2009 report gave a little more detail. Of possible
note, while not included in the chapter's key messages summary, the 2009 report's
main text did refer to the possibility of thresholds being crossed whereby an ecosystem
stops changing gradually but instead moves quickly to a new state. Climate-related
thresholds are also referred to elsewhere in the report, such as in relation to health
issues.
The 2001 USGCRP report did not just look at possible US climate and ecosystem
change but also how its population and economy would develop throughout the
21st century. It predicted that the population would increase by between just 20%
and almost a doubling, depending on various emigration scenarios. Employment
forecasts were also variable, from barely rising at all through to rising up to 2070