|
Titel |
Supporting transparency and uncertainty analysis of derived hydrological data through development of common exchange formats |
VerfasserIn |
Peter Taylor, Paul Sheahan, Doug Palmer, Stuart Hamilton, Matt Fry, Michael Natschke, David Blodgett, David Briar, David Valentine, Simon Cox |
Konferenz |
EGU General Assembly 2013
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 15 (2013) |
Datensatznummer |
250077551
|
|
|
|
Zusammenfassung |
Many derived hydrological data sets, such as river discharge, are based
on relationships (ratings) built up over time through observations of related
phenomenon, such as river level. The derived data have often been treated as
actual observations neglecting the underlying procedure used to develop the
relationship1.
There is an increased interest in exposing the underlying assumptions in derived hydrological
data2, and more generally
in environmental data3,
with a view to more consistent estimation and understanding of the uncertainty in such
data.
It has been argued4
that the lack of available information on rating relationships, and the associated observational
data, is one of the major contributors to the problem of transparency of hydrological models.
Members of the joint WMO/OGC Hydrology Domain Working Group are developing an
exchange format, WaterML2.0 part 2, to address exchange of rating tables, gaugings
(observations) and cross-section information with a view to enhancing transparency of
derived hydrological data. Improving the availability and definition of the rating
relationships, and related metadata, will improve transparency, and support the estimation of
uncertainty.
The work thus far has focused on harmonizing structures for representing rating tables of
sufficient resolution to allow linear interpolation between points. This choice was based on
the complexity in the number of curve representations within the domain and across data
holders. A reusable data-structure has been developed to express information critical to
understanding uncertainty by representing metadata relating to measurement sites and
phenomenon ranges. For example, changes in controlling channel bottom features over a
range of river levels, or expressing upper and lower limits for extrapolation of a rating curve.
The group is continuing to explore the essential information to include in data exchange to
allow users to assess the uncertainty associated with rating curves; working to strike a
balance between capturing actual assessments of uncertainty and/or quality and providing
the necessary information that allows experts to perform their own assessment. |
|
|
|
|
|