 |
| Titel |
Stochastic downscaling of numerically simulated spatial rain and cloud fields using a transient multifractal approach |
| VerfasserIn |
M. Nogueira, A. P. Barros, P. M. Miranda |
| Konferenz |
EGU General Assembly 2012
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 14 (2012) |
| Datensatznummer |
250063666
|
|
|
|
|
|
| Zusammenfassung |
| Atmospheric fields can be extremely variable over wide ranges of spatial scales, with a scale
ratio of 109-1010 between largest (planetary) and smallest (viscous dissipation) scale.
Furthermore atmospheric fields with strong variability over wide ranges in scale most likely
should not be artificially split apart into large and small scales, as in reality there is no scale
separation between resolved and unresolved motions. Usually the effects of the unresolved
scales are modeled by a deterministic bulk formula representing an ensemble of
incoherent subgrid processes on the resolved flow. This is a pragmatic approach
to the problem and not the complete solution to it. These models are expected to
underrepresent the small-scale spatial variability of both dynamical and scalar fields due to
implicit and explicit numerical diffusion as well as physically based subgrid scale
turbulent mixing, resulting in smoother and less intermittent fields as compared to
observations. Thus, a fundamental change in the way we formulate our models is
required.
Stochastic approaches equipped with a possible realization of subgrid processes and
potentially coupled to the resolved scales over the range of significant scale interactions range
provide one alternative to address the problem. Stochastic multifractal models based on the
cascade phenomenology of the atmosphere and its governing equations in particular are the
focus of this research. Previous results have shown that rain and cloud fields resulting from
both idealized and realistic numerical simulations display multifractal behavior in the
resolved scales. This result is observed even in the absence of scaling in the initial
conditions or terrain forcing, suggesting that multiscaling is a general property
of the nonlinear solutions of the Navier-Stokes equations governing atmospheric
dynamics. Our results also show that the corresponding multiscaling parameters
for rain and cloud fields exhibit complex nonlinear behavior depending on large
scale parameters such as terrain forcing and mean atmospheric conditions at each
location, particularly mean wind speed and moist stability. A particularly robust
behavior found is the transition of the multiscaling parameters between stable and
unstable cases, which has a clear physical correspondence to the transition from
stratiform to organized (banded) convective regime. Thus multifractal diagnostics
of moist processes are fundamentally transient and should provide a physically
robust basis for the downscaling and sub-grid scale parameterizations of moist
processes.
Here, we investigate the possibility of using a simplified computationally efficient
multifractal downscaling methodology based on turbulent cascades to produce statistically
consistent fields at scales higher than the ones resolved by the model. Specifically, we are
interested in producing rainfall and cloud fields at spatial resolutions necessary for effective
flash flood and earth flows forecasting. The results are examined by comparing downscaled
field against observations, and tendency error budgets are used to diagnose the evolution
of transient errors in the numerical model prediction which can be attributed to
aliasing. |
|
|
|
|
|
|