 |
| Titel |
Analyzing scales of precipitation in general circulation models and observations |
| VerfasserIn |
Nicholas Klingaman, Gill Martin, Aurel Moise |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250149205
|
| Publikation (Nr.) |
 EGU/EGU2017-13536.pdf |
|
|
|
|
|
| Zusammenfassung |
| General circulation models (GCMs) have been criticized for failing to represent observed
scales of precipitation, particularly in the tropics where simulated rainfall is often said to be
too light, frequent and persistent. Previous assessments have used temporally or spatially
averaged precipitation, which offers little actionable information for model developers, since
the physics-dynamics interactions that produce precipitation occur at the native gridscale and
timestep.
We introduce a set of diagnostics (ASoP1) to compare the spatial and temporal scales of
precipitation across GCMs and observations, which can be applied to data ranging from the
gridscale and timestep to regional and sub-monthly averages (Klingaman et al., 2017). When
applied to data from ten GCMs, ASoP1 diagnostics reveal that far from the “dreary”
persistent light rain implied by daily mean data, on the native timestep and gridscale most
GCMs produce a broad range of intensities (1–100 mm day−1). Several GCMs, including
the Met Office Unified Model (MetUM), show quasi-random behaviour that may alter the
spectrum of atmospheric waves. Averaging to a common spatial (≈ 600 km) or temporal
(3 hr) resolution reduces inter-model variability, demonstrating that averaging masks
intrinsic model behavior.
To further explore intermittent MetUM timestep precipitation and connect it to longer-
and larger-scale biases, we analyze MetUM simulations at a range of horizontal resolutions,
including ≈ 16 km simulations with and without a deep convective parameterization (Martin
et al., 2017). With parameterized convection, intermittency is largely insensitive to resolution
and timestep length, as are larger- and longer-scale variability. Switching off the
parameterization results in very persistent but very sporadic rainfall. On the ≈ 100 km scale,
the spectra of oceanic 3-hr and daily mean rainfall in the parameterized configurations agree
well with satellite-derived rainfall estimates. At ≈ 10-day scales, the spectra indicate a lack
of intra-seasonal variability. Over tropical land, MetUM often underestimates daily mean
rainfall (related to a poor representation of the diurnal cycle) and lacks sub-seasonal
variability.
References
Klingaman NP, Martin GM and Moise AF, 2017: ASoP (v1.0): A set of methods for
analyzing scales of precipitation in general circulation models. Geosci. Model Dev., 10,
57–83, doi:10.5194/gmd-10-57-2017.
Martin GM, Klingaman NP and Moise AF, 2017: Connecting spatial and temporal scales
of tropical precipitation in observations and the MetUM-GA6. Geosci. Model Dev., 10,
105–126, doi:10.5194/gmd-10-105-2017. |
|
|
|
|
|
|