Outcomes of synthesis - Runoff Prediction in Ungauged Basins: Synthesis across Processes, Places and Scales - page 382

will happen next (Blöschl, 2006 ); and (ii) understanding
that will enable extrapolation to new situations (Kumar,
2011 ). Hydrological synthesis is a vehicle for creating
these connections and it is hoped that this topic contributes
to this kind of synthesis.
In the spirit of Bronowski ( 1956 ), synthesis must lead to
the discovery (indeed, in a deep sense, creation) of order in
what otherwise appears as disorder. In this sense the syn-
thesis that we have carried out has indeed led to the
discovery of order, as outlined above. By organising the
synthesis along processes, places and scales we have man-
aged to gain deep insights that otherwise would have
remained hidden. Synthesis across scales has revealed the
inter-connection between various signatures, including
how seasonality is a connective tissue underlying all of
the signatures. Synthesis across places has revealed the
critical role that aridity and seasonality play in catchment
responses. Synthesis across scales has revealed the strong
dependence of the area

Newtonian (error propagation) and Darwinian (compara-
tive) approaches to uncertainty estimation.
The outcomes of the synthesis presented here represent
an accumulation of knowledge, i.e., distilled knowledge.
This accumulation of knowledge has only been possible
through an organised community effort, a distillation
achieved through synthesis organised across processes,
places and scales ( Figure 12.16 ). Given that both the
PUB initiative in general and its synthesis in this topic
have, in each case, been the result of community efforts,
what lessons have been learned from this effort? How can
the community organise itself in the future to benefit from
and build upon the progress made so far?

12.4.2 Role of the community
Accumulation of knowledge is an important and legitimate
way to advance the state of a science, especially when it is
an applied science such as hydrology. This topic has
clearly demonstrated the value of comparative assessment
of previous experiences with prediction methods to con-
tribute to an accumulation of knowledge. Accumulation of
knowledge requires the community to organise, and there
are important lessons that can be learned from the experi-
ence of putting together this topic. From the results
reported in several case studies ( Chapter 11 ) we have
noticed differences in the way hydrological communities
within developed countries have organised and responded
to the challenges of PUB, in contrast to the situation in
developing countries. Lack of organisational and institu-
tional drivers can prevent the accumulation of local know-
ledge, and can compound the disadvantages that
hydrologists and practitioners in many developing coun-
tries already face, in terms of complexity of the hydrology
and the lack of data and research funding.
In spite of the apparent success of the comparative
assessment of model performances, it must be said that
our assessment was hampered by the lack of critical and
basic information that many published modelling studies
failed to report; this is due to the absence of a reporting
protocol and the lack of organisation across the modelling
community. Indeed, as noted by Gupta et al.( 2008 ):

-

time scale dependence on various
signatures, and on prediction performance. Most import-
antly, the study has revealed what methods work best in a
particular climate. This is a particularly new and useful
result that will advance hydrological prediction generally.
It is hoped that this topic contributes to converting data on
predictions in ungauged basins into information and into
knowledge that the community can use as a whole (see
Figure 12.15 ).
Going further, the effort has led to a higher level of
synthesis of two competing paradigms that the hydrology
community has been grappling with: Newtonian and Dar-
winian. We have outlined the need for such a synthesis for
improved predictions as well as for the advancement of the
science, and have illustrated it with examples. This has also
led to the call for a new uncertainty framework that is
better suited to PUB, that combines the advantages of the

As a community, we have fallen into reliance on measures and
procedures for model performance evaluation that say little more
than how good or bad the model-to-data comparison is in some
“ average ” sense.

In the past, reporting of most modelling studies was more
heavily focused on demonstrating that the chosen models
worked well and not so much on the underlying reasons,
and reports especially failed to provide the hydrological
insight needed to interpret

the modelling results. For

Figure 12.15. From data to information to knowledge to understanding
in predictions in ungauged basins. presentationload.com. knowledge to accumulate it is essential that the