Status of the CWB-NCAR Project

 

Task

Title

Name

Status

Date

1

Task #1 – Support for the WRFVar component of the CWB operational system and preparation for the upgrade of WRFVar 3.0

Guo,

M-H Chen,

Eric Chiang,

Cheryl

 

 

 

 

 

 

 

1.1

Support for the WRFVar component of CWB operational system:

 

-       Guo continues to respond to questions coming from CWB staff, particularly Eric.

-       Guo also provides support to Maxine Chen

-       This is on-going support work throughout the year

-       A major goal this year is to ensure the operation of WRF-Var (Version 2.1) at CWB with CV=5, 6-h cycling

 

Guo

On-going

 

1.2

Preparation for running WRFVar 3.0 in CWB

Guo, Eric

 

 

1.2.1

Install and compile the WRFVar 3.0 code on CWB IBM

-       Eric downloaded WRF 3.0 and WRF-Var 3.0, and they are working for the CWB 45 km domain.

Guo, Eric

done

 

1.2.2

Test the WRFVar 3.0 with the exact same configuration used in WRFVar2.1

-       For this test, we need to identify a case for testing

-       Suggest comparing the June 2007 case for these two versions.

-       This will be primarily a CWB task, with NCAR support

Eric, Guo

To be done

 

1.2.3

Test the accessory functions of WRFVar 3.0

[Gen-BE, error tuning utilityÉ]

-       This is primarily a CWB task. CWB staff should test options/functions of interest, and let NCAR staff know, if there are problems.

-       NCAR staff will trouble shoot.

Eric, Cheryl, Guo

To be done

 

1.2.4

Complete the retrospective 3-hour cycling run tests with WRFVar 3.0

-       Maxine Chen found that CWB and NCAR Version 2.2 produce different results. Currently, Guo and Maxine are performing checks.

-       An important job is to perform 3-h cycling in preparation for the 27/9/3 km domains.

-       This should be done by CWB staff with NCAR support.

-       Should have a teleconference on this topics.

-       Suggest that we also perform the test on June 2007 case.

 

Eric, Cheryl, Guo, Maxine Chen

To be done

 

1.3

Develop the capability of the ground-based GPS ZTD data assimilation

 

This task will be primarily carried out by Guo. He will begin the work soon. Ya-Ting can perform some tests at CWB before coming to NCAR in July. This should be discussed in the next teleconference.

Guo, Braun

 

 

1.3.1

Development and implementation of forward, tangent linear, and adjoint code for ZTD assimilation

 

Guo, Ya-Ting

TBD

 

1.3.2

Testing the capability of assimilating CWB GPS ZTD data

Guo, Ya-Ting

TBD

 

1.3.3

Comparing the results with the GPS PW data assimilation

Ya-Ting, Guo

TBD

 

1.3.4

Preparation of a report to summarize the results

Guo

TBD

 

 

 

 

 

 

2

Task #2 – Preparation of the WRF/DART ensemble adjustment filter data assimilation system for semi-operational testing at CWB

 

 

 

 

 

 

 

 

2.1

Upgrade to WRF 3.0

Jeff Anderson, Nancy Collins

done

 

2.2

Implement forward observation operator for satellite retrieved thickness into WRF/DART

-       Coding is finished for the thickness operator, and surface observations operator.

-       Hui Liu will provide the code to Y-T Lin for testing

Hui Liu

Essentially done

 

2.3

Improve the computing efficiency of WRF/DART and perform one-month assimilation and compare with 3DVAR

-       Most of the work is done

-       Hui Liu will work with Y-T Lin for the one-month testing

-       Need to discuss specification for the one-month testing. In particular, we need to coordinate the work with WRF-Var, so that we can compare the performance between WRF/DART and WRF-Var over a one-month period.

-       Also, we need to assess the impact of: (a) COSMIC GPS RO data, (b) 3-h versus 6-h cycling.

-       Should have a teleconference on this.

Hui Liu

On-going

 

2.4

Assimilation and ensemble forecast for typhoon Krosa and Wipha of 2007

-       Y.-T. Lin has done the Krosa case. Y-T Lin should provide the results (via .ppt file) to NCAR for review.

-       Some model runs blew up for Y-T LinŐs testing at CWB. Need to work with Hui Liu

-       Hui will begin the Wipha case soon

Hui Liu, Y-T Lin

On-going

 

 

 

 

 

 

3

Near Real-Time Data Collection and Display of GPS Precipitable Water Vapor Estimates in Support of TIMREX

 

 

 

 

 

 

 

 

3.1

Real-time data collection

-       Currently, CWB is producing 2 hr data set, and data is delivered after 1 hr. The processing begins 5 min after collection, and takes another 30 min to finish. Therefore, the total data latency is 3 hr and 35 min. This is not optimal (for mesoscale application). It would be desirable to shorten the latency.

-       CWB has begun testing LAPDOG, and will try to produce 1 hr data set, and to speed-up data delivery.

-       We have met the contract requirements. But, John will continue to provide support.

Braun

Essentially done. Some support

 

3.2

Real-time data analysis

-       Data Processing software is running on a machine at CWB

-       Completed 35 min after data delivery

-       The data processing can benefit from increased memory. CWB will purchase and increase memory. This can reduce the processing time from 35 min to 20 min.

Braun

Done

 

3.3

Web Display

-       An off-line version has been developed.

-       Need to port to CWB and turn it online.

-       Currently, we are waiting for a CWB machine (should have a teleconference) to be available. Also, the web page will evolve after receiving user feedback.

Maggie, Braun

Mostly done. Additional updates

 

3.4

Interpolating surface observations

-       John Braun rwrote the code

-       John may modify the code to detect outliers

-       Code is working on CWB machine

-       J.-S. Hong and Ya-Ting are running PW assimilation with WRF 3D-Var using data produced from this system

Braun,

Essentially done

 

3.5

Post Processing:

-       Running operationally at CWB

-       Currently, it is using the rapid orbit from IGS. It can also be run with IGS final orbit, which will require more waiting time.

-       Currently, post processing is run one-day after real-time.

-       DonŐt have enough compute power for post-processing. This can benefit from a better machine at CWB.

-       should have a teleconference to discuss this task.

 

Braun

Essentially done

 

3.6

Technical Support

 Braun

On-going

 

 

 

 

 

 

4

Task #4 – Continued interaction on WRF data assimilation systems

 

 

 

4.1

Update and improve the CWB project web pages on both the CWB and the UCAR sides

-       Guo will develop a wiki at NCAR.

-       All project members can have write access to the wiki to report status of their task

Guo

On-Going

 

4.2

Site visit to CWB

-       Kuo may visit CWB on 3rd week of May

-       Need to decide dates for the final view and decide who should attend from NCAR.

-       Need to set the dates for Cheryl Terng and J-S HongŐs visit to NCAR. The objectives are to review the various tasks, and to discuss next yearŐs plan. We expect this to take place in the  September/October 2008 time frame.

-       Check whether Ben Jou will visit NCAR (because Ben contribute to the GPS work). (Bill K. will talk with Ben)

 

 

 

 

 

 

 

 

5

Task #5 – Support the installation and testing of the NCAR High-Resolution Land Data Assimilation System (HRLDAS)

 

 

 

5.1

Lectures and training

-       Fei Chen visited CWB for two days, and gave a seminar on land surface modeling

-       4 day tutorial by Michael to 8-15 people, including CWB, university students and scientists.

-       Michael prepared one month and a half forcing data for testing, for the period of June 15 – July 31 2007

Fei Chen, Michael B.

Done

 

5.2

Implementation, testing, and evaluation of the offline HRLDAS:

-       J.-S. Hong wants to try on his own. Therefore, CWB does not want NCAR staff to do all the work.

-       Need to work on Different data types. Still working with them on what are available. This includes gridded precipitation data from CWB, 8-km data (60S-60N) CMORPH precipitant data, and solar radiation data.

-       This is done on all three model domains (45/15/5).

-       Need a teleconference to review status. If necessary, NCAR staff may step in and perform the task as required.

 

Michael B., J.-S. Hong,

Vivian  Tsay

On-going, continued interaction with J.-S. Hong

 

5.2.1

NCAR scientists will provide suggestions for the implementation of the HRLDAS in CWB's IBM machine by CWB staff

-       We donŐt know whether the code is working on the CWB IBM.

-       The code is running on CWB linux machine

-       There are security issues related to CWBŐs IBM

-       should have teleconference to review this.

Micharl Barlage

On-going

 

5.2.2

Test the HRLDAS in the CWB environment using NCAR's data. NCAR will run HRLDAS at NCAR for a period of one week, to evaluate it and to make sure that the simulation of HRLDAS at CWB for the same time period produces the same results as NCAR HRLDAS simulations

-       NCAR side of the work has been done.

-       One month and a half month data have been prepared

-       Need to know the status of the IBM, and CWBŐs results,  so that we can compare

Michael Barlage, J.-S. Hong

On-going

 

5.3

Ingest Taiwan local data in HRLDAS and perform an evaluation:

 

-       This is a task primarily to be carried out by CWB staff

-       Need to know the status of this task at CWB

J.-S. Hong

On-going

 

5.4

Provide consultant/technical support to CWB staff in their performing an offline HRLDAS experiment:

 

-       Need to know the status of this work at CWB

-       Michael to provide support as needed

J.-S. Hong

On-going

 

5.4.1

NCAR will provide suggestions about the data needed to evaluate HRLDAS results:

 

-       CWB got flux tower data

-       Need to know what data are available to CWB for assimilation

Michael B.

On-going

 

5.4.2

NCAR will help with an evaluation conducted by CWB to compare HRLDAS results with available observations, and provide suggestions to adjust HRLDAS parameters according to the evaluation results

-       Need to review status

-       Identify availability of MODIS data products and local land cover data

Michael Barlage, J.-S. Hong

On-going

 

 

-       J.-S. Hong will be visiting NCAR, some time in September 2008

-       Should discuss the work plan and J.-S. HongŐs visit

-       We need a teleconference to coordinate TASK #5 soon