The 29th BMB-NCAR Project Meeting

 

Time: 6:00 pm~6:45 pm, 17 July, 2006, Room 2072

NCAR/MMM Participants: Dr Huang, Dr Guo, Dr Wang and Min Chen

IUM Participants: Dir Wang, Dir Zhang, Shui-yong Fan and Jiqin Zhong

 

1. Dr Huang had a short introduction at the beginning of the meeting. According to previous two meetings, a detailed working plan of BMB-NCAR project has been made and all participants of this project can start to work following the plan.  Task 4 is completed with the report of the implementation of WRFVAR provided by Shui-yong. Dr Huang and Min will be the contact points of NCAR and BMB, respectively. Each project participant should send a biweekly report on the working to both Dr Huang and Min, by every second Friday. Dr Huang and Min will write a summary by the following Monday. (Hans added after the meeting: The next reporting time is July 28.)

 

2. Dr Huang made some comments on the unreasonable cold bias case. 1) From the results provided by IUM, the initial conditions differ a lot between MM5 and WRF, which is unreasonable; 2) Dr Wang needs a tar file containing the initial and boundary condition data to check the problem; 3) We should try Dr GuoÕs suggestion first, i.e. using identical IB condition for both WRF and MM5 model. 4) Try Dr WangÕs suggestion, i.e. use WPS to replace WRFSI for initialization. 5) Physics tuning should be the last task.

 

3. Jiqin simply introduced the verification results of the quasi-operational WRF temperature forecast. The forecasted 2-m temperature at the grid point closest to Beijing Observatory (54511) is used to compare with that of 54511 for verification. The average difference is around -3~-5 degree. The worst situation is -8 degree.

 

4. Dir Zhang mentioned that the IB conditions of the nested domain are directly interpolated from the 27 km domain for the WRF model. As a consequence, the nested surface information from the coarse domain may not be compatible with the high resolution. Dr Wang mentioned that some nesting boundary bugs were just fixed in the latest WRF Version 2.1.2+. So we have to make sure that nesting setup is correct first.

 

5. Dir Wang pointed out that the most urgent job currently is the real-time assimilation of all available non-conventional data for WRF model, especially for the coming FDP test of this summer. She inquired the possibility of using a single 9 km domain with AVN background and WRFVAR analysis for un-conventional observations for operation. Dr Wang, Dr Guo and Dr Huang replied that there should be no technical problem for running a single 9km domain for operation. However, the domain should be large enough (or the same domain size as 27 km) to avoid any trouble caused by lateral boundaries. To run WRF-Var on a new domain needs a recomputed BE. Another option is to keep the current 9 km size, but the forecast time should be shortened to 24 hours due to the validation of the boundary data. Dr Huang suggested using the MM5 9 km BE for WRFVAR temporarily.

 

6. Dr Wang also made several suggestions to resolve the cold bias problem: 1) Delay the physics tuning; 2) Update the model version to WRF Version2.1.2 which has the nesting bug fixes; 3) Check the nesting setup and related namelist; 4) Replace WRFSI by the new WPS. 5) Try the 9 km single domain, if the above methods cannot solve the problem.

 

7. A task list:

1)    IUM staff will provide a tar file containing SI, WRFVAR, real namelists and the IB conditions.

2)    Dr Wang will check the setup or repeat the run if necessary.

3)    IUM staff will repeat the WRF experiment using the IB conditions identical to the MM5 IB conditions.

4)    Preparing for the single 9 km domain?

 

 

Notes added by Hans:

After the meeting, Bill and Hans had a short discussion. They came to another suggestion: run an experiment with the current operational system (WRF-Var + MM5) plus PW. This is a safer way to bring PW into operation in a short time.