Benefit of the MTG candidate Infra-Red Sounding mission to
regional forecast
1. Project Objective
The objective of the study is to document the added value of water
vapor observations derived from a hyperspectral infrared sounding instrument on
a geostationary satellite for regional forecasting. An Observation System
Simulation Experiment (OSSE) is proposed to document the impact of such
observations. In such an experiment, synthetic observations, either retrieved
water vapor profiles or simulated radiance observations, are assimilated into a
regional forecast model to analyze the impact of such observations on the
performance of the forecast.
2. Detailed Work Plan
Three phases are planned.
2.1 Phase I
2.1.1 Preparation
A period of 5 days for the
OSSE will be identified and proposed to EUMETSAT, including a justification of
the particular choice.
A
short description of the nature model (MM5), the forecast model (WRF) and the
data assimilation scheme (WRF-Var) will be written. Necessary tests of the two
models will be performed.
2.1.2 Nature run
A nature run will be carried out using the nature model. The model
is planned to have a resolution of 4 km and 500x500x35 grid points. The output
will be at every 10 min. The results will be made available to EUMETSAT.
2.1.3 Initial OSSE
The
modeled temperature and humidity fields extracted from the nature run will be
assimilated to give initial states for forecast runs. Forecasts from these
initial states will be compared to the results of the nature run. This OSSE
will document the ideal case, namely that observations by IRS can be used to
derive the exact state of the atmosphere. Uncertainties of observation errors
will be considered in this OSSE.
The
second OSSE run will be based on temperature and humidity profiles generated by
retrieval simulations using the results of the nature run. These retrieval
profiles will be made available by EUMETSAT. This OSSE shall document a
semi-realistic case, namely where uncertainties in the retrieved profiles due to
retrieval uncertainties are introduced. The assimilation scheme will use the
observation errors generated by the retrieval method.
2.2 Phase II
Additional
analyses of the Phase I results will be performed and when required additional
OSSE runs will be executed.
In
addition to the OSSEs mentioned in Phase I, an OSSE based on the retrieved
humidity profiles will be performed to document a semi-realistic case, namely
where uncertainties in the retrieved profiles due to retrieval uncertainties
are introduced. The assimilation scheme adopted for the forecast model shall
use the observation errors generated by the retrieval method. The results of
this OSSE will be used to document the importance of humidity information
derived from MTG-IRS observations.
2.3 Phase III
The
calibration of the OSSE setup will be performed using radiosonde observations.
First,
pseudo radiosonde observations will be extracted form the nature run. Then
three experiments will be conducted.
Exp
1. No-obs. The forecast model will be run over the 5-day period.
Exp
2. Pseudo-obs. A virtual reality run will be made using the data assimilation
scheme, the pseudo observations and the forecast model.
Exp
3. Real-obs. A real reality run will then be made using the virtual reality
configuration, except the real radiosonde observations will be used.
The
impact of the pseudo observations (Exp 2 – Exp 1) and that of the real
observations (Exp 3 – Exp 1) will be assessed by using forecast error
reductions. The results will be used to evaluate the OSSE results for the
MTG-IRS candidate mission.
3. Time table
|
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
|
Phase I |
|
|
|
|
|
|
|
|
|
|
|
|
|
Phase II |
|
|
|
|
|
|
|
|
|
|
|
|
|
Phase III |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
KO |
|
PR |
|
|
|
|
|
MR |
|
|
FR |
KO: Kick-off meeting, 19 January 2007, NCAR
PR: Preliminary results Review, 5-6 March 2007, EUMETSAT
MR: Mid-term Review,
September 2007, EUMETSAT
FR: Final Review, December 2007, NCAR
4. The Team
The work will be carried out at NCAR and the University of Hawaii.
Dr Xiang-Yu Huang, a member of the MMT, will lead the project. He
will work on the project for a total of 2 months time. He is currently working
at NCAR and has the following corresponding address:
Dr Xiang-Yu Huang
NCAR, MMM, P.O. Box 3000, Boulder, CO 80307-3000, USA
Email: huangx@ucar.edu
Tel: 1-303-497-8975
Hongli Wang, a PhD student, will work full time on the project.
Dr N.N., a project scientist, will work on the project part time.
Dr Xin Zhang, University of Hawaii, will collaborate with the
team. Two visits are also planned.