Environmental Engineering Reference
In-Depth Information
by the direction and geographical shape of the change in the observed
and simulated variables, not by the absolute values of trends and spatial
anomalies. The best established and robust results in detection and attri-
bution of anthropogenic influences are those that have been documented
for temperature increases at global and continental scales, at the surface,
in the troposphere, and in the oceans. When moving from temperature to
other parameters and from global and continental scales to smaller regional
scales, the confidence and robustness of the results diminish but are still
high for many other measures of physical changes (e.g., sea level pressure,
temperature extremes, zonally averaged precipitation, atmospheric moisture
and surface specific humidity). The evaluation of the IPCC has also included
detection and attribution of changes that have not been analyzed through
the statistical machinery of formal D&A but are either closely related to
changes that have been so analyzed, or have been proven consistent in
a qualitative way with experiments including anthropogenic forcings and
not consistent with model simulations forced by natural inputs alone. In all
cases the scientific reasoning and understanding of the mechanisms linking
anthropogenic forcings to these changes buttress the evidence from obser-
vational records and models.
In Table 1.1 we list the physical parameters of the climate system
whose changes have been detected and attributed. We list first those D&A
results that have been documented in IPCC AR4, along with a measure of
confidence assigned, followed by a list compiled from more recent peer-
reviewed literature.
Formal D&A, as just described, relies on model simulations, by which a
treatment (human factors included) versus control experiment (only natural
factors included) is set up for D&A of climate variables, but D&A of im-
pacts requires an additional modeling step, by which the behavior of the
climate system is translated into effects on the natural of human systems
under study. The additional modeling adds limitations and uncertainties,
but, where it is possible using current understanding, D&A of impacts is
methodologically similar to D&A of physical parameters' change, and it is
only the degree of significance or uncertainty in the results that will have
to account for the compounding of modeling steps and the approximation
errors thus added. For many types of impacts, however, a direct modeling,
with the ability to include the confounding factors that are often required
in addition to the climate changes, is not possible, and D&A has to follow
a multi-step pathway:
