|
Titel |
Alternative spatial configurations to reflect landscape structure in a hydrological model: SUMMA applications to the Reynolds Creek Watershed and the Columbia River Basin |
VerfasserIn |
Bart Nijssen, Martyn Clark, Naoki Mizukami, Oriana Chegwidden |
Konferenz |
EGU General Assembly 2016
|
Medientyp |
Artikel
|
Sprache |
en
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 18 (2016) |
Datensatznummer |
250130195
|
Publikation (Nr.) |
EGU/EGU2016-10413.pdf |
|
|
|
Zusammenfassung |
Most existing hydrological models use a fixed representation of landscape structure. For
example, high-resolution, spatially-distributed models may use grid cells that exchange
moisture through the saturated subsurface or may divide the landscape into hydrologic
response units that only exchange moisture through surface channels. Alternatively, many
regional models represent the landscape through coarse elements that do not model any
moisture exchange between these model elements. These spatial organizations are often
represented at a low-level in the model code and its data structures, which makes it difficult to
evaluate different landscape representations using the same hydrological model. Instead, such
experimentation requires the use of multiple, different hydrological models, which in turn
complicates the analysis, because differences in model outcomes are no longer constrained by
differing spatial representations. This inflexibility in the representation of landscape
structure also limits a model’s capability for scaling local processes to regional
outcomes.
In this study, we used the Structure for Unifying Multiple Modeling Alternatives
(SUMMA) to evaluate different model spatial configurations to represent landscape structure
and to evaluate scaling behavior. SUMMA can represent the moisture exchange between
arbitrarily shaped landscape elements in a number of different ways, while using the same
model parameterizations for vertical fluxes. This allows us to isolate the effects of changes in
landscape representations on modeled hydrological fluxes and states. We examine the effects
of spatial configuration in Reynolds Creek, Idaho, USA, which is a research watershed with
gaged areas from 1-20 km2. We then use the same modeling system to evaluate scaling
behavior in simulated hydrological fluxes in the Columbia River Basin, Pacific Northwest,
USA. This basin drains more than 500,000 km2 and includes the Reynolds Creek
Watershed. |
|
|
|
|
|