Environmental Engineering Reference
In-Depth Information
decades, centuries, and millennia. It discusses best estimates and uncertain-
ties, including those relating to feedbacks (such as carbon cycle feedbacks
whereby more carbon is emitted from the biosphere in a warming planet).
Chapters 4 through 6 comprise the core of this report. They present
climate changes and impacts responses as a function of warming, that is,
corresponding to stabilization targets of 350 to 2,000 ppmv CO
2
-equivalent,
considering time frames up to 2100, as well as very long time frames several
thousand years in the future.
1.2 ATTRIBUTION
We offer here a brief summary of detection and attribution results as
they provide part of the foundation of the discussion of future projected
changes in the physical climate system and the impacts on natural and hu-
man systems that could ensue from them. This summary is not intended to
be comprehensive of all detection and attribution studies to date, but will
start from the IPCC Earth Assessment Report (AR4), updating it by those more
recent results that are specifically relevant to this report.
Formal detection and attribution of an anthropogenic influence over the
physical climate system
is based on analysis of spatial and temporal patterns
in observations of climate parameters and on comparison of their statisti-
cal characteristics with those of the same patterns as simulated by climate
models. Because models can be integrated by applying the known external
forcings in designed experiments (natural only, anthropogenic only, natural
and anthropogenic jointly
1
) or in unforced mode (i.e., a control simula-
tion), the behavior of the system subjected to different forcings as well as in
control mode can be characterized, and the observed behavior of the real
climate system can be compared to test consistency with a naturally vary-
ing process or with a process subjected to externally (especially manmade)
forcings, to a given degree of statistical confidence. The progress of formal
detection and attribution (D&A) is thus linked inextricably to the accumula-
tion over time and space of quality observations that allows computation
of robust statistics of the parameters under study and comparison to the
ability of climate models to reliably simulate those parameters' natural and
forced behavior. Importantly, formal D&A compares
spatial and/or tempo-
ral patterns
of change in observations and model simulations, not simply
magnitudes of changes, seeking to test the consistency of the process-driven
behaviors between models and observations. These behaviors are defined
1
Each of these examples can be split into more complex designs with single natural/anthro-
pogenic forcings administered in isolation or jointly.
