263 / 280
[
] 263
primary species-occurrence records (both specimens
and observations) from some 1000 databases and 200
data providers around the world, and covering a
diverse range of taxa and ecosystems. A high propor-
tion of these records are geo-referenced, and ongoing
efforts in the data providing communities promote the
necessity and value of providing an accurate geo-loca-
tion for records. The GBIF virtual database represents
a unique resource for Earth Observation studies which
require ground-truthing data, whether historical (to
study change over time) or contemporary. GBIF
exposes the data through a web site and several web
services.
2. A climatological data provider, such as the
National Center for Atmospheric Research (NCAR)
Geographic Information System (GIS) portal provides
web access to free, global datasets of climate change
scenarios. These data (spanning 50 years from 2000
to 2050) have been generated for the 4th Assessment
Report of the IPCC by the Community Climate System
Model (CCSM). Climate models are an imperfect
representation of the earth’s climate system and
climate modellers employ a technique called ensem-
bling to capture the range of possible climate states.
A climate model run ensemble consists of two or more
climate model runs made with the exact same climate
model, using the exact same boundary forcings, where
the only difference between the runs is the initial
conditions. The NCAR portal provides several climate
change scenarios. Of these, the constant 20th century
scenario shows the least increase in future surface
temperature, the B1 and A1B scenarios display moder-
ate increases and the A2 scenario results in the largest
used, GEOSS infrastructure will enhance access to high quality
earth observation datasets of relevance to ENM.
Step four
– Determine which historical and future scenario
climatological data are needed for ENM of the selected group of
organisms. This may include measurement of how the species in
question has/have responded to recent climate changes, thus
strengthening any inferences related to how spatial ENMs will
change temporally.
8
Step five
– Determine which modelling algorithms will most
accurately and precisely predict shifts in distribution and abun-
dance for the selected group of organisms. Identify the reporting
needs in terms of data accuracy and error propagation.
Step six
– Download the selected species occurrence data (eg
from GBIF) and environmental and climate data (eg from IPCC)
to the modelling workbench.
Step seven
– Run the models and present the outputs as a series
of maps and predicted species’ ranges or abundances, as appro-
priate. Measure uncertainty in model outputs under the range of
desired scenarios so a realistic depiction of policy options is avail-
able to policy-makers. This approach resembles that of the IPCC
in presenting different climate change scenarios depending on vari-
ations in emission reduction efforts.
This scenario is but one example of a broad-scale application
for biodiversity data. Biodiversity is also affected by other factors
such as tropical deforestation, for which other scenarios can be
produced. However, these additional scenarios also build on the
same pool of primary biodiversity data as the described climate
change scenario.
The overall prediction system architecture comprises five main
components from the biodiversity and climate change Societal
Benefit Areas, as follows:
1. A biodiversity data provider, such as the GBIF Data Portal
(http://data.gbif.org) provides unified access to some 130 million
Tests using the common roadside skipper butterfly
A) The
Amblyscirtes vialis
distribution projected for the year 2000; B) The
Amblyscirtes vialis
distribution projected for the year 2050 under the IPCC B1
climate change scenario. Light markers corresponds to 100 per cent of presence probability; grey markers to 50 per cent of probability
Source: Nativi et al. 2007b.
Amblyscirtes vialis
photo by Erik Nielsen
A
B
S
OCIETAL
B
ENEFIT
A
REAS
– B
IODIVERSITY