ALADIN/HIRLAM data assimilation training
See here for original. Requires hirlam.org account.
Practical info
- Location: The Hungarian Meteorological Service Headquarter (OMSZ)
- Date: 10-15 February 2019
Background reading
- Data assimilation concept and methods, Bouttier and Courtier, 1999
- Introduction to data assimilation, ECMWF material
- Overview of satellite observations and their role in NWP, ECMWF material
Presenters
| Name | |
|---|---|
| AB | Antonin Bucanek |
| AT | Alena Trojakova |
| BS | Benedikt Strajnar |
| DS | Daniel Santos |
| EW | Eoin Whelan |
| PM | Paulo Meideros |
| RH | Rafiq Hamdi |
| RR | Roger Randriamampianina |
| RS | Roel Stappers |
| TA | Trygve Aspelien |
Program
Monday
| Time | Presenter | Presentation |
|---|---|---|
| 0900 - 0945 | Eoin Whelan | Observations for NWP[EW1] |
| 0945 - 1030 | Alena Trojakova | Bator and ODB[AT1] |
| 1100 - 1130 | Daniel Santos | ECMWF working Environment |
| 1300 - 1330 | EW & AT | Introduction to the practicals |
| 1330 - 1530 | Eoin Whelan | BUFR observations (solutions) |
| 1600 - 1730 | AT & EW | Bator (solutions) |
Tuesday
| Time | Presenter | Presentation |
|---|---|---|
| 0900 - 0945 | Roger Randriamampianina | Observation Screening[RR1] |
| 0945 - 1030 | RR & TA | Bias correction of observations Obsmon[TA1] |
| 1100 - 1130 | Daniel Santos | Harmonie System part I |
| 1300 - 1530 | RR & AT & TA & PM | Screening and Monitoring |
| 1600 - 1730 | AB & RR | Screening (ALARO) practicals |
Wednesday
| Time | Presenter | Presentation |
|---|---|---|
| 0900 - 0945 | Roel Stappers | Variational Data assimilation[RS] |
| 0945 - 1030 | Antonin Bucanek | Minimization & B Matrix computation[AB] |
| 1100 - 1130 | Daniel Santos | Harmonie system Part II |
| 1330 - 1530 | Antonin Bucanek | Minimization and Single Obs |
| 1600 - 1730 | Antonin Bucanek | DA exercise |
Thursday
| Time | Presenter | Presentation |
|---|---|---|
| 0900 - 0945 | Rafiq Hamdi | Surface DA[RH] |
| 0945 - 1030 | Trygve Aspelien | Surface DA II[TA] |
| 1100 - 1130 | Eoin Whelan | Cycling - Input/Output/Archiving in Harmonie |
Friday
| Time | Presenter | Presentation |
|---|---|---|
| 0900 - 0945 | Alena Trojakova | Initialisation[AT2] |
| 0945 - 1030 | Benedikt Strajnar | Diagnostics in DA[BS] |
| 1330 - 1730 | AT & BS & RR | Diagnostic Tools |
- EW1A summary of observations used in NWP. Conventional (GTS), satellite (EUMETSAT), radar (OPERA) and other observations will discussed in the context of the ALADIN/HIRLAM System.
- AT1An overview of BATOR program to convert various observation data into ODB-1 format, to perform a blacklisting, to assign observation errors for conventional data and to make a preselection (geographic,timewindow, channel and sensors) of data. A short review of ODB and ODB applications (odbsq, MANDALAY, ODBTOOLS). |
- RR1How the good observations are selected for data assimilation through quality control, blacklisting and data thinning. What do we do with known biased observations?
- TA1Obsmon is a tool used to monitor the use of observations from the ODBs used in 3/4D-Var and CANARI. It consists of two parts, the first extract information from ODB and the second part is the visualization done in a browser.
- RSTheoretical background for variational data assimiliation (Minimization algorithms, preconditioning, Large scale constraint)
- ABAn overview of minimization configuration (inputs, environment, namelist parameters) and checking of its execution. Various methods of B matrix computation: spinup, EDA (femars a festat programs).
- RHTheoretical and scientific background for surface assimilation in NWP.
- TAThis second part will focus more on practical aspects on how to do surface assimilation with the external surface model SURFEX and how this links to the NWP model and HARMONIE script system
- AT2A short introduction to initialization (DFI, IDFI, space consistent coupling).
- BSIntroduction of various diagnostic techniques for DA: Analysis of residuals, Degrees of Freedom for Signal, Moist Total Energy norm, covariances of residuals and tuning of background/observation errors and thinning based on a-posterior diagnostics.