 |
| Titel |
Evaluation of ice and snow content in the global NWP model GME with CloudSat |
| VerfasserIn |
Sonja Eikenberg, Kristina Fröhlich, Axel Seifert, Susanne Crewell, Mario Mech |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250053192
|
|
|
|
|
|
| Zusammenfassung |
| Ice clouds have a large impact on the Earth’s climate system due to their effects on
the global radiation budget. A good description of ice clouds is therefore a major
challenge for both climate and NWP models. The CloudSat Cloud Profiling Radar
offers the so far unique opportunity to vertically resolve clouds from space – in
contrast to the numerous passive satellite-based sensors. Due to its high resolution and
the near-global coverage it is predestined for the evaluation of global models and
finally offers an observational constraint for the water constituent ice. Especially for
NWP models, which are subject to ongoing enhancements, the development of
a continuous evaluation technique is of major interest. First steps in the design
of such a technique are made within the evaluation of a novel ice microphysical
parameterization of the global NWP model GME of the German Weather Service
DWD.
Since the radar reflectivity factor is not a direct model parameter, two principal evaluation
approaches are possible: Observation-to-model and model-to-observation. We undertake
both approaches, with the goal of exploiting the full informational content and
determining the pros and cons of each approach. For the observation-to-model
approach the ice water content (IWC) retrieval (Austin et al., 2009) is utilized, for
the model-to-observation approach the radar simulator QuickBeam (Haynes et
al., 2007) is applied. The first approach is easy to compute and the actual model
parameters are compared. However, the uncertainties of this approach are not easily
assessed; three parameters of the particle size distribution (PSD) are retrieved from one
measurement and several assumptions are included. The second approach to some extent
avoids the problem of the retrieval uncertainties and is closer to the actual physics
by simulating the reflectivity factor from the model forecast of the gaseous and
hydrometeor component, including the model’s assumptions on PSDs . However, ice
crystals are modelled as soft spheres in QuickBeam, and not as the actual particle
habit.
To improve the comparability between model and observations, we develop criteria
which are applied to each matching pixel: (1) only temperatures lower than -10°C to
avoid liquid and part of mixed phase, (2) top of convection below 1 km height to
reduce sub-grid and mixed phase effects, (3) cloud cover larger than 50 % to ensure
homogeneous conditions, and (4) total column attenuation not larger than 3 dBz to avoid
attenuation. Because CloudSat cannot distinguish between cloud ice and snow,
model IWC also includes both, with snow contributing on average 85 % of total
mass.
The evaluation shows that in comparison to CloudSat the new prognostic precipitation
scheme of GME performs better than the old diagnostic scheme concerning the magnitude of
IWC and reflectivity factors and the shape of the frequency distributions. As a consequence,
the new scheme became operational on 2 February 2010. Furthermore, the applied evaluation
technique enables the assessment of processes within clouds, benefiting from the vertically
resolved CloudSat data, and thereby the identification of possible sources of error in the
model parameterizations. For example, the results indicate an overestimation of ice water
path (IWP) in the new GME version, which might be due to a too long residence time of
snow in the air. Changes in snow fall speed successfully reduce IWP by up to a
factor of two. This additional improvement became operational on 1 December
2010.
References:
Austin, R.T., Heymsfield, A.J., and Stephens, G.L. (2009) Retrieval of ice cloud
parameterizations using the CloudSat millimeter-wave radar and temperature. J. Geophys.
Res. 114.
Haynes, J.M., Marchand, R.T., Luo, Z., Bodas-Salcedo, A., and Stephens, G.L. (2005) A
multipurpose radar simulator package: QuickBeam. Bull. Am. Met. Soc., 131, 1997-2017. |
|
|
|
|
|
|