Rationale for NEON Domains,
letter from Bruce Hayden, NNDC, to COREO on 8 September 2005
NEON Response to Colleagues on Climate Domain Boundaries
The NNDC would like to thank COREO for their efforts to organize the responses
to the climate domain map. It provided the needed materials to evaluate the
requests for changes.
First, recall that the NEON domains are in fact climate domains. Only climate
data was used in their development. Our two-fold approach to defining climatic
domains was:
1. Used cluster analysis of climate state variables. (Hargrove, 2004)
2. Used resultant wind vectors to delineate climatic regions based on airmass seasonality. (Bryson, 1966; Mitchell, 1969)
We explored a number of classification options to arrive at the climate
domains. We used as many as 29 variables to derive 100 zones to 9 variables
to derive
25 potential zones. After significant debate we settled on 9 variables
to derive 25 potential zones. These zones, at one km resolution, were
then subjected
to a 25 km smoothing.
Of the 25 potential zones in the lower 48 states, and with guidance of
the maps of airmass seasonality, we discussed and arrived at 16 of
the 20 final
domains.
The zones in Alaska and the tropics were climatically distinct bringing
the total number of domains to 20.
The 20 domains were voted on by the NNDC and unanimously approved.
• Number of days above 90°F during the local growing season
• Number of days below 32°F during the local non-growing season
• Precipitation sum during the local growing season
• Precipitation sum during the local non-growing season
• Number of days with measurable precipitation during the local growing season
• Number of days with measurable precipitation during the local non-growing season
• Soil plant-available water holding capacity to 1.5 m
• Total solar insolation during the local growing season, including clouds, aerosols, slope and aspect physiography• Total solar insolation during the local non-growing season, including clouds, aerosols, slope and aspect physiography
• Wind vectors
• Streamline and seasonal front analysis from the winds
Why isn’t there a node in Antarctica? The NNDC made this decision based on the design of the NEON science plan, which calls for an urban-to-wild gradient in every NEON node. Antarctica is not included in NEON because it lacks urban–to-wild gradient component as required by the NEON science plan.
The final map with 20 domains, whose boundaries
were placed using the approach outlined above,
was accepted
by the NNDC
and unanimously approved.
In the coming years, consortia of scientists
will become the stewards of NEON climate domain
research facilities.
We recognize that other
designs and other
input data would result in different boundaries
that might be more attractive to some COREO
groups. NEON
by necessity,
is being designed
to be a continental
platform and the approach described here takes
into account variation at
a continental scale.
The NEON climate domain map represents a reproducible
map of climatically derived, statistically
similar land units
that, in total, capture the
climatic variations
of the United States. It is envisioned that
the map will be used by a broad range of
NEON stakeholders
as they establish
domain-specific
and
cross-domain
partnerships.
The specific siting of infrastructure within
domains
will be determined though an as yet to be
specified peer-review
mechanism.
The NNDC addressed the issues associated
with requests for alterations in the domain
boundaries.
After a great
deal
of discussion, we reaffirmed
(by
unanimous
vote) the concept of climatic domains,
our method of arriving at them, and the map now
on the NEON
webpage.
Again, thank you for your help and your
support. The work ahead will not end
soon.
For the NNDC
Bruce