Nonhydrostatic Numerical Model Study on Tropical

1
Non-hydrostatic Numerical Model Study on
Tropical Mesoscale System During SCOUT DARWIN
Campaign
Wuhu Feng1 and M.P. Chipperfield1 IAS, School of
Earth and Environment, University of Leeds, U.K
1. INTRODUCTION The SCOUT-O3 tropical aircraft
campaign took place in Darwin Australia in
November 2005 and obtained some detailed
information of chemical species including water
vapor and tracers which can be used to study the
transport in the TTL region. However, the global
chemical transport model (CTM) TOMCA/SLIMCAT can
not reproduce the observed H2O because that it
does not consider the transport of H2O from
troposphere into the lower stratosphere and also
partly due to the accuracy and resolution of
meteorological conditions. In this poster, we
show preliminary some results from mesoscale
simulations with the Weather Research and
Forecasting model (WRF), which can ultimately be
used to force TOMCAT.

• 2. Weather Research and Forecasting (WRF) Model
• Next generation mesoscale NWP system.
• Equations Fully compressible, Euler
  non-hydrostatic.
• Vertical coordinate Terrain-following
  hydrostatic pressure.
• Horizontal grid Arakawa C-grid staggering.
• Time Integration Time-split integration using a
  3rd order Runge-Kutta scheme with smaller time
  step for acoustic and gravity-wave modes.
• Detailed microphysics (microphysics, PBL,
  radiation, cumulus parameterisations and surface
  physics etc).
• http//www.wrf-model.org

Table 1. WRF Model Configuration
Fig. 1. An example Comparison of SLIMCAT with M55
data on 25 Nov 2005.
Run D01
Fig. 2. IR cloud images from GMS-5 satellite on
16 Nov 2005. Also shown are M55 and Falcon
aircraft locations.
Fig. 3. Observed 24-hour accumulated total
precipitation (mm) since 16 Nov 2005.
Run D02
Fig. 7. 700 hPa wind at 0800 UTC 16 Nov 2005 from
D02. Low level jet brings a large amount of water
vapour to north Australia.
Fig. 5. Simulated 24-hour accumulated total
precipitation (mm) from WRF since 16 Nov 2005
from two WRF simulations.
Fig. 4. Domains used in WRF model simulations and
the corresponding terrain height (m).
Fig. 8. As Fig. 7 but for diagnosed equivalent
potential temperature (?e) at 850 hPa. The
distribution of high ?e is consistent with
rainfall belt.
Fig. 9. Evolution of cloud water (04UTC, 07UTC,
09UTC) during the first scientific flight on 16
Nov. 2005.

• Results
• The coarse resolution WRF run (D01)
  underestimates the observed accumulated total
  precipitation (Figure 5 and Figure 3). The higher
  resolution run (D02) improves on this. However,
  the simulated location of rainfall is still not
  satisfactory. Further sensitivity experiments
  need to be performed.
• Low-level jet brings a large amount of water
  vapour to north Australia and the distribution of
  high ?e (warm and high humidity) is consistent
  with the distribution of rainfall belt (Figures
  3, 7 and 8). There is intensive divergence on the
  upper levels near Darwin.
• The modelled top of cloud water around Darwin
  was below 4Km (Figures 9) while the cloud top
  reached 12 Km around Indonesia during the first
  flight .
• Quality of ECMWF Temperature need to be checked
  (Figures 10).

Fig. 6. 100hPa streamline and divergence at 0900
UTC 16 Nov 2005 from D01.

• Conclusions
• Mesoscale WRF model provides high spatial and
  temporal resolution simulations which will be
  useful to force our SLIMCAT/TOMCAT CTM.
• Detailed comparison with observations measured
  during Darwin Campaign especially the transport
  of water vapor in the TTL will be studied.

Fig. 10. Same As Fig. 1, but for the first
scientific flight on 16 Nov. 2005.
Acknowledgements. We are grateful for the use of
SCOUT-O3 campaign data and to NCEP for the global
analyses. This work was supported by the EU.

References Stamarock, W. C., et al., A
description of the Advanced Research WRF version
2, 2005.
For more information about this poster Wuhu
Feng, School of Earth and Environment, University
of Leeds, Leeds, LS2 9JT Email
fengwh_at_env.leeds.ac.uk