Certainty clouds the IPCC by lindayy

VIEWS: 8 PAGES: 3

More Info
									 Certainty
 clouds the IPCC
                                          Sinclair Davidson & Alex Robson

     ‘When I use a word,’ Humpty
     Dumpty said, in a rather scornful
     tone, ‘it means just what I choose
                                          ‘I  t’s almost certain’, The Weekend Aus-
                                              tralian headline told us, that human
                                          activity is the cause of global warming.
                                                                                            ing authors it is a difficult effort) will no-
                                                                                            tice the terms ‘very likely’, ‘very high con-
                                                                                            fidence’ and the like. But what do these
     it to mean—neither more nor less’.   The Intergovernmental Panel on Climate            terms mean? The IPCC employs very
     ‘The question is,’ said Alice,       Change (IPCC) brought out a draft ver-            clumsy terminology and, indeed, man-
     ‘whether you can make words mean     sion of its ‘Summary for Policymakers’ in         ages to conceal precise meanings in its
     so many different things’.           February this year. The data within the           report. In this article we investigate what
     ‘The question is,’ said Humpty
                                          report is final, but the report itself is still   the IPCC is telling us and show that it is
     Dumpty, ‘which is to be master—
                                          subject to ‘editorial adjustments to fig-         far from clear that global climate change
     that’s all’.
                                          ures’. Importantly, the summary states            is induced by humans.
                                          that there is a ‘very high confidence that
     —Lewis Carroll                       the globally averaged net effect of hu-           The IPCC Process
     Through the Looking-Glass            man activities since 1750 has been one            The IPCC was created in 1988 by two
                                          of warming’. In other words, human                United Nations organisations: the UN
                                          events since the Industrial Revolution            Environmental Programme and the
 Sinclair Davidson is Professor in        began have contributed to global climate          World Meteorological Organisation.
 the School of Economics, Finance and     change. Most of the increase in average           The IPCC exists to ‘to assess on a com-
 Marketing, RMIT University.              temperature is very likely due to human           prehensive, objective, open and transpar-
 Alex Robson is a Lecturer in Economics   activity. Readers who struggle through            ent basis the scientific, technical and
 at the Australian National University.   the entire 21-page document (with 33              socio-economic information relevant to
                                          drafting authors and 18 draft contribut-



MARCH 2007                                                      7                                               R E V I E W
understanding the scientific basis of risk
of human-induced climate change, its             Figure 1: Temperature Anomaly (1850–2006)
potential impacts and options for adap-
tation and mitigation’ (our emphasis). In
this regard, the IPCC is in the process of         0.6
releasing its ‘Fourth Assessment Report            0.4
on Climate Change’. These assessment               0.2
reports, written by experts in the field, are
comprehensive summaries and syntheses                0
of the extant peer-reviewed literature            -0.2
(non-peer-reviewed literature can be in-          -0.4
cluded, but must be justified). The draft
                                                  -0.6
assessment reports are meant to provide
a ‘balanced and complete assessment’ of           -0.8



                                                         850

                                                                860

                                                                       870

                                                                                  880

                                                                                         890

                                                                                                900

                                                                                                       90

                                                                                                              920

                                                                                                                     930

                                                                                                                            940

                                                                                                                                   950

                                                                                                                                          960

                                                                                                                                                 970

                                                                                                                                                        980

                                                                                                                                                               990

                                                                                                                                                                      2000
                                                                                                                                                                      2005
the body of knowledge. There are then
two review processes—first by ‘experts’
and second by ‘government and experts’.           Source: http://www.bom.gov.au/web01/ncc/www/cli_chg/timeseries/global_t/0112/
The review process is meant to be ‘ob-            global/latest.txt
jective, open and transparent’. What is
important to note is that experts in the
field are summarising work they have
contributed to, and experts whose own            Figure 2: Global Mean Temperature (1850–2006)
work is being summarised are reviewing
those summaries. In other words, there is           30
no external disinterested evaluation. The
                                                    25
IPCC process is incestuous. This, how-
ever, is not as problematic as one might            20
imagine. The original research is peer-re-
                                                    5
viewed and is less likely to be corrupted.
What is problematic is the interpretation           0
of the work contained in the summaries
                                                     5
for policymakers.
                                                     0
                                                         850

                                                                860

                                                                       870

                                                                                  880

                                                                                         890

                                                                                                900

                                                                                                       90

                                                                                                              920

                                                                                                                     930

                                                                                                                            940

                                                                                                                                   950

                                                                                                                                          960

                                                                                                                                                 970

                                                                                                                                                        980

                                                                                                                                                               990

                                                                                                                                                                      2000
                                                                                                                                                                      2005
What we’re being told
The IPCC has concluded that global
warming is occurring. This would seem            Source: Derived from the Australian Bureau of Meteorology Data
to be uncontroversial. Many people have          http://www.bom.gov.au/web01/ncc/www/cli_chg/timeseries/global_t/0112/global/lat-
seen the diagram shown in Figure One.            est.txt
This shows the difference between annual
global mean temperature and the average              People looking at a contrived figure                              such as ‘extremely unlikely’, ‘highly like-
global mean temperature for the period          such as our Figure 1 can be forgiven for                               ly’ and ‘very high confidence’ have abso-
1961 to 1990. Relative to the average, it       thinking that global warming is a huge                                 lutely no scientific basis, and there is no
can easily be seen that the global tempera-     urgent problem. There is no reason why                                 scientific justification whatsoever for as-
ture has risen quite dramatically in the last   the temperature is shown relative to an                                sociating numerical, model-based proba-
30 years. The IPCC report contains a fig-       average over the period 1961 to 1990.                                  bilistic assessments (if that is indeed what
ure very much like our Figure 1, showing        There is nothing obviously special about                               they are) with such loaded expressions. A
dramatic global warming.                        those years. What the IPCC does not                                    5 per cent chance means precisely what
      Looking at the data, the warmest year     show is the actual global mean tempera-                                it says, whereas ‘highly unlikely’ could
was 1998 and the equal coldest years were       ture over the period 1850 to 2006. We                                  mean anything. It needlessly confuses the
1909 and 1911. The temperature differ-          show that data in Figure 2.                                            reader.
ential between the warmest and the cold-             The IPCC then also says that this                                       Hidden away in the footnotes, the
est years is 1.09 degrees centigrade. The       warming is ‘extremely unlikely’ to be due                              IPCC report translates these terms. ‘Ex-
range of maximum and minimum global             to natural variation in climatic processes,                            tremely unlikely’ means less than 5 per
mean temperature over the past 156 years        and that it has ‘very high confidence’ that                            cent probability and ‘very high con-
is just over 1 degree centigrade.               this is due to human behaviour. Terms                                  fidence’ means ‘at least a 9 out of ten


R E V I E W                                                                   8                                                                                       MARCH 2007
  chance of being correct’, or 90 per cent
  probability. These terms indicate ‘the
  assessed likelihood, using expert judge-
                                                     Table 1: IPCC Expert Assessment of Human Impact on Extreme
  ment, of an outcome or a result’. The lay-         Weather Events
  man reading this might think that there
  is a 90 per cent probability that human                                                            Likelihood
  activity is causing global warming—in-                                                             of a human
                                                                                                                           Probability
  deed, the Australian newspapers reported                                                         contribution to
  the IPCC in that manner. Yet, it is un-                                                          observed trend
  clear what that 90 per cent represents.             Warmer and fewer cold days and
        The IPCC helpfully provides a table                                                              Likely           More than 66%
                                                      nights over most land areas
  to guide our understanding of climatic
  trends and human effects on those trends.           Warmer and more frequent hot
  We reproduce part of that table, and aug-           days and nights over most land                Likely (Nights)       More than 66%
                                                      areas
  ment it with the statistical meaning that
  the IPCC has hidden in its footnotes. Of            Warm spells / heat waves.
  the seven extreme weather events iden-              Frequency increases over most land          More likely than not    More than 50%
  tified by the IPCC, two have a better               areas
  than 66 per cent probability of human               Heavy precipitation events.
  contribution, and five have a better than           Frequency (or proportion of total
                                                                                                  More likely than not    More than 50%
  50 per cent probability of human contri-            rainfall from heavy falls) increases
  bution. Yet, the IPCC also tells us that            over most areas
  the probability of human contribution               Area affected by droughts increases         More likely than not    More than 50%
  to climate change is better than 90 per
                                                      Intense tropical cyclone activity
  cent. The whole, according to the IPCC,                                                         More likely than not    More than 50%
                                                      increases
  is substantially greater than the sum of its
  parts.                                              Increased incidence of extreme
                                                                                                  More likely than not    More than 50%
                                                      high sea level (excludes tsunamis)
  How big is big?
                                                     Source: adapted from IPCC table Summary for Policy Makers, February 2007
  Scientific method requires researchers to
  posit a hypothesis, collect data, and then
  test the hypothesis. Hypothesis-testing
  is a well-known technique taught in all          readers, with large data sets, confidence        duces a guess as to what it all means. That
  first-year statistics classes. In other words,   levels have to be increased, so a 90 per         may well be an informed guess but it is
  hypotheses are falsifiable. That human           cent confidence level would not then be          not a scientific process. How then does
  activity is causing the planet to warm is        valid—the hypothesis is falsified.               the IPCC form its opinions? What pro-
  a testable hypothesis. The IPCC seems to              It is not clear from the report wheth-      portion of scientists are guessing more
  imply that it has tested that hypothesis         er the IPCC has, in fact, undertaken such        than 90 per cent, and what proportion
  and failed to reject it. It is at least 90 per   an analysis. It is more likely that it has       are guessing less than 90 per cent? What
  cent confident that the hypothesis is not        neither a testable model nor data avail-         is the non-scientific public meant to be-
  false. Should we be concerned? Probably,         able for external researchers to replicate       lieve? After all, humans are either causing
  not. If this is what the IPCC has done, it       such a test. In other words, the IPCC’s 90       global warming, or they are not. Despite
  has very weak evidence. Hypothesis-test-         per cent confidence level has not emerged        years of work, and commitments to ob-
  ing normally requires confidence levels          from a scientific process; it has emerged        jectivity, the IPCC is unable to provide
  either of 90 per cent, 95 per cent, or 99        from scientists evaluating whether they          the answer to that question. At present, it
  per cent. Ninety per cent is the weakest         think their own work is correct. Further,        expects us to take its word at face value—
  level of confidence. Yet, the IPCC tells         in contrast to the IPCC’s own require-           again, a notion that is not scientific.
  us that its understanding is based ‘upon         ment, the 90 per cent confidence level is
  large amounts of new and more compre-            not open, not objective, and not trans-                                             I PA
  hensive data, more sophisticated analysis        parent.
  of data, improvements in understand-                  The IPCC purports to undertake a
  ing of processes and their simulation in         huge task summarising the extant litera-
  models, and more extensive exploration           ture on global warming and the extent to
  of uncertainty ranges’. As any elemen-           which humans contribute to that warm-
  tary textbook of statistics reminds its          ing. Yet, when all is said and done, it pro-


MARCH 2007                                                             9                                                 R E V I E W

								
To top