ASIAN REGIONAL ACTIVITIES Overview of Asian THORPEX Regional Plans and Activities (Submitted by Tetsuo Nakazawa)

Transcription

1 WORLD METEOROLOGICAL ORGANIZATION COMMISSION FOR ATMOSPHERIC SCIENCES INTERNATIONAL CORE STEERING COMMITTEE FOR THORPEX Eighth Session Offenbach, Germany (2 4 November 2009) CAS/ICSC-8/DOCX.X (XX X.2009) Item: X.X ASIAN REGIONAL ACTIVITIES Overview of Asian THORPEX Regional Plans and Activities (Submitted by Tetsuo Nakazawa) After the last ICSC-7 in November 2008, there were several activities under the ARC. These were: 2 nd Asian THORPEX Scientific Workshop 6 th Asian Regional Committee Meeting 1. 6 th ARC meeting The 6 th ARC meeting was held, just after the 2 nd Asian THORPEX Scientific Workshop, in Hangzhou, February At the meeting there were members from China(2), Japan(2), Korea(2), Russia(1) and many observers. A member from India (Dr. Bohra) could not come due to his urgent business. For the report on the 2 nd Asian THORPEX Scientific Workshop, please refer to the Status Report from China. The objectives in this meeting were to exchange information on the recent progresses of the national THORPEX activities, especially the results of T-PARC in 2008 to implement the Asian THORPEX activities in the WWRP Strategic Plan to elect Chair and co-chairs 1.1 Adoption of the minutes at the last ARC meeting After the chair explained the minutes of the 5 th ARC meeting, the members adopted the minutes. 1.2 The Chair Report Before going to the main agenda, the chair addressed a short note on the ARC activities after the last ARC meeting. He introduced a new member after the 5 th ARC meeting from Korea, Dr. Dong-Eon Chang, who took over Dr. Hee-Sang Lee, who was suddenly dispatched to the educational institute for one year from January The chair also mentioned that the T-PARC was successfully conducted in collaboration among ARC, NARC and ERC and also coordinated with the TCS-08 program. 1.3 Status Reports There were four status reports from member countries. (China) Dr. Gong presented the current status of the model development (GRAPES) and the development of the TIGGE Global Ensemble Prediction System. He also showed the result of the Beijing Olympic 2008 Forecast Demonstration Project and Research Development Project at CMA. Dr. Zong added the observational plan, which is closely related with the THORPEX activities in China. (Korea) Dr. Chang presented the status on the KEOP Project. He showed the internal KMA webpage of all typhoon track forecasts, written in Cyclone XML. He also showed that the KMA is now working for the OSE of Typhoons, in collaboration to the Universities, such as Yonsei Univ.

2 CAS/ICSC-8/Doc.X.X, p. 2 For T-PARC, he expressed that some targeted observation results may have a chance of degrade if you use dropsondes. (Japan) Dr. Nagata presented the current status of the model development in JMA. (Russia) Dr. Oleg Pokrofsky reported on the optimization of the upper sounding network in Russia. 1.4 WWRP Strategic Plan and Post TPARC activities The members discussed the WWRP Strategic Plan and post T-PARC activities. There were several comments on this. 1.5 Election of Chair and Co-Chairs In the Terms of Reference (TOR) of ARC, the terms of Chair and co-chairs are both 2 years, up to 4 years. Dr. Nakazawa served 4 years ( 2 terms with 2 years apart ). Dr. Chen also served 4 years. However, it was discussed that the first term was tentative after a sudden passed away of Dr. Nakamura. Thus the first term should not be counted. Although the current members from China are Drs. Dehui Chen and Jishan Xue, China proposed to change members to Drs. Jiandong Gong and Donghai Wang. Also China proposed to add a honorable membership in TOR. If he or she served at some certain periods and did a great contribution to the ARC activities, then he or she has a right to be selected as a honorable member. 1.6 Action Items The 6 th ARC meeting agreed the following action items. ARC-5/1 ARC chair to summarize the comments for the WWRP Strategic Plan to send to the THORPEX IPO ARC-5/2 ARC chair to prepare a new Terms of Reference for the ARC, to modify the membership ARC-5/3 A member of Korea will inform to the ARC members on the next ARC meeting information 1.7 On the next ARC meeting Dr. Chang kindly proposed the next meeting in Korea, but the date and venue will be informed to members later.

3 CAS/ICSC-8/Doc.X.X, p Status Report from China 2.1 Progress on CMA TIGGE Data Archive Center Serving as the TIGGE data archive center, CMA setup the TIGGE data archive platform which is a Linux cluster and has LDM and MARS deployed for collecting TIGGE data from ECMWF and NCAR and archiving all the collected data at CMA locally. To provide TIGGE data access service, CMA developed the TIGGE data portal and put it into running in It is built upon J2EE architecture and supports the services of TIGGE data retrieval and data request/reply. And, with the supports of the BRIDGE project, the service of accessing the distributed TIGGE data based on the Grid middleware technologies is being developed by ECMWF, CMA and CAS Network Information Center (CNIC), and will be demonstrated in the near future. 2.2 The Second Asian THORPEX Science Symposium The Second Asian THORPEX Science Symposium was held in Hangzhou, China, from 18 to 20 February Four sciences themes were discussed in the meeting: (1) Predictability and Dynamical Processes associated with tropical cyclone, monsoonal severe rainfall, MJO, sand and dust storm. (2) Data assimilation and observing systems, and improvement in NWP model and targeting technique. (3) TIGGE relative research and archive center activities (4) Observation field campaigns and others There were 84 scientific abstracts submitted to the science committee of the meeting and 70 were accepted to communicate in the symposium. This meeting is the largest Asian THORPEX regional scientific meeting by far and 10 scientists were invited to make special report, besides 20 oral and 30 poster presentations. Near 100 researchers attended the meeting and the participants came from Japan, Korea, Russia, France and China. There were several highlights in the symposium: T-PARC is the largest and successful regional observation experiment by far and Dr. Tetsuo Nakazawa s report drew great attention; China has made great progress in TIGGE archive data center construction and TIGGE data application; Numerical model progress in Japan; Development of China s GRAPES model. 2.3 B08FDP/RDP The fourth B08FDP/RDP workshop was held in Guangzhou, China on April B0RDP stands for Beijing 2008 Olympics Mesoscale Ensemble Prediction Research and Development Project ( B08RDP ), which is designed to contribute to the better understanding and improved forecasting of high impact weather events in summer season in Beijing from 6 to 36 hours with mesoscale ensemble prediction system (MEPS) in support of decision-making in 2008 Olympic meteorological service. In that workshop, activities and progress since last workshop was reviewed and discussed, and the main issues are: a product processing and distribution platform was completed, which dealt with the product transmission, encoding/decoding, product generation and display; all B08RDP participants (NMC/CMA, CAMS/CMA, MRI/JMA, NCEP, ZAMG, EC) have made significant progress in the project and in other relevant fields: Initial perturbation technique, perturbed initial & lateral boundary conditions, model perturbation, bias correction and combination, high resolution simulation experiments, etc. B08RDP highlighted its support for Olympic Games in real time, and the EPS products (including special products for specific Olympic venues) from all six participants were transmitted to local forecasters and proved great successful in Olympic weather support. Statistical and synoptic verification demonstrate the importance of

4 CAS/ICSC-8/Doc.X.X, p. 4 MEPS for the prediction of high impact weather (HIW) and indicate the advantages of MEPS over global ensemble prediction system(geps) for the predictions of surface variables as well as advantages of ensemble forecasts versus operational deterministic forecasts. B08RDP lasted for 5 years, with 3-4 years of research and development, data transfer test, MEPS establishment, 1-2 years for demonstration, analysis of results and inter-comparisons. The implementation of B08RDP made substantive achievements and gained rich experiences. 2.4 Progress of project The study of key application technology for TIGGEbased ensemble prediction system In 2007, A research project of national public sector, "The study of key application technologies for TIGGE-based ensemble prediction system", was granted by Chinese Meteorological Administration (CMA). While solving key techniques for applications of TIGGE products, this project also tries to practice the concept of GIFS advocating by THORPEX plan and build a prototype system of interactive forecast. By studies in 2 years, the project team gains a series of results: having updated the NMC/CMA T213 global ensemble forecasts in CMA, having initially established a test platform for global/regional coupled ensemble prediction based on Chinese new generation numerical model GRAPES, having developed the coupled-model of the meteorological models GRAPES and hydrological XXT model, having initially established a reservoir user-oriented conceptual model for interactive forecast analysis application. During research, a very difficult problem in TIGGE data applications is how to found key techniques for interactive forecast system based on the hydrological user information. For example, the user service for great mass TIGGE data, the downscaling application techniques of TIGGE data for hydrological users, the synthesis techniques for multi-model EPS data, and the combining ways for user information to establish the interactive forecast system etc. The first phase of this project will be finished in December But, how to set up a prototype of GIFS by user-oriented and practice the interactive forecast concepts based on multi-model TIGGE ensemble prediction still need to be further studied. 3. Status Report from Japan 3.1 T-PARC 2008 Special Observations by JMA THORPEX Pacific Asian Regional Campaign (T-PARC) was conducted in 2008, under the international partnership, to understand the life cycle of tropical cyclones over the western North Pacific, from genesis through intensification to recurvature and extra-tropical transition. The Japan Meteorological Agency (JMA) performed special observations and sensitivity analysis for tropical cyclones (TCs) as a part of T-PARC. The special observations were conducted, aiming for the improvement of the numerical weather prediction (NWP) performance in cases for TCs. The observational data included dropsonde deployments by manned aircraft, additional radiosondes by research vessels and observation sites, and the MTSAT rapid-scan operations. The data was distributed through the Global Telecommunication System, so that the data could be used in operational numerical weather prediction systems around the world. 3.2Sensitivity Analysis Sensitivity analysis was performed using a singular vector (SV) method to provide guidance for the special targeted observations. To make the plan for the special targeted observations, the information of sensitivity area at the time of the observation was needed in advance. Thus, the sensitivity analysis was performed for the forecast fields (T+24h and T +48h) of operational Global Spectral Model (GSM) at a TL959L60 resolution. In addition to daily sensitivity analysis for three fixed target areas (named as GUAM, TAIWAN and JAPAN), JMA conducted another sensitivity analysis for an adaptive target area (named as MVTY) in the vicinity of a TC location because an adaptive target area is preferable when

5 CAS/ICSC-8/Doc.X.X, p. 5 targeting a TC although a fixed target area is useful for inter-comparison of the sensitive area calculated by each provider (Figure 1). The MVTY target area was automatically defined according to the TC position forecasted by the operational GSM. The optimization time intervals (OTI) of SVs were chosen to be 48 hours for GUAM, TAIWAN and JAPAN regions and 24 hours for MVTY region. The norm used for measuring the amplitude of perturbations at both initial and final time is based on moist total energy (moist TE) as described in Ehrendofer et al. (1999). The SVs calculated for all target areas were moist SVs, because moist processes including both large-scale condensation and deep convection were implemented in the tangent linear and adjoint model at a T63L40 resolution for the global domain. 3.3OSE for Sinlaku and Jangmi The major concern in Japan is the impact of targeted observation for the tropical cyclone track forecast. Several OSE experiments have been performed to evaluate the impact for Sinlaku and Jangmi. The result indicates that the overall improvement in the track forecast was identified. For Sinlaku case, the track forecast was improved by percent in 0 12 hour before recurvature periods and by 10 percent in hour after recurvature periods. However, for Jangmi both tracks with or without special sonde observations are identical, suggesting that the impact is small when the forecast is good. The intercomparison of the track forecast among several centers; JMA, KMA, ECMWF and NCEP, is now on-going to see the common features and differences in centers. 4. Status Report from Korea 4.1Post THORPEX Pacific Asian Regional Campaign (T-PARC) activities 4.1.1Adaptive observation guidance From 1 August to 4 October 2008, T-PARC has been performed in the western North Pacific to investigate structures, targeting, and extratropical transition of tropical cyclones. To suggest airborne targeting regions, we have provided the total energy singular vector (TESV) guidance based on the fifth generation mesoscale model (MM5) to European Center for Medium-Range Weather Forecasts (ECMWF) Data Targeting System and Japan Meteorological Agency (JMA) T-PARC websites. For T-PARC typhoons, MM5 TESV guidance was well consistent with SV guidance from other institutions (e.g., ECMWF, Naval Research Laboratory, and JMA) Observation system experiments The impact of the additional dropsonde observations during T-PARC to the typhoon forecasts is investigated for typhoon SINLAKU (200813) and JANGMI (200815). The Weather Research and Forecasting (WRF) model and corresponding three dimensional variational (3DVAR) data assimilation system and Ensemble Kalman Filter (EnKF) are used to investigate the impact of additional dropsonde observations. Overall, the targeted dropsonde data has positive impact on the typhoon track and intensity forecast with some variations depending on the experimental framework (e.g., type of data assimilation system, etc.). 4.2Special Observation Campaign of High Impact Weathers (HIWs) associated with ProbeX 4.2.1Ensemble-based sensitivity analysis for summer climate in Korea To identify summer climatology of sensitivity feature of Korea, ensemble sensitivity analysis technique had been applied. Climatological sensitivity pattern of various forecast variables (e.g., sea level pressure, 500 hpa height, and 850 hpa temperature) during August

6 CAS/ICSC-8/Doc.X.X, p is investigated using the advanced research version of WRF model. The regions of high composite sensitivity can inform that additional observation affects the forecast. A large fraction of the frequently sensitive region is observed on the maritime of Korea. So, data assimilation for the strong forecast sensitivity area, west and south sea of Korea, is performed in order to characterize the potential impact of the observation. These results demonstrate that predictability of summer weather will be improved, if additional observation over strong sensitive region is performed. In this regard the special observations over the maritime of Korea are planned from 23 August to 10 September Special observation campaign 2009 In order to improve and identify predictability of catastrophic weather phenomena, the intensive observation campaign was conducted in the maritime of Korea. The uppersounding observations on the shipboards (The Gisang 2000 and the Eardo) were performed at every 6-hour interval from 24 August to 7 September 2009 at the West and South Sea of Korea. At Ulleung-island in East Sea and super-site of Haenam in South Coast, 6-hourly radiosonde soundings from 24 August to 10 September 2009 were also conducted. Besides in situ intensive observation, special observations on the rain and cloud microphysics at Ulleung-island were provided by both remote sensing (e.g., radiometer, GPS instruments, X- band dual-polarization, Vertically Pointing x-band radar (VertiX), Micro-rain radar (MRR), and ceilometers, etc.) and in-situ instruments (e.g., AWS, radiosonde sounding at every 12- hour interval, etc.). We plan to use the WRF 3DVAR systems to assimilate observed data during ProbeX 2009 and investigate their sensitivity to predictability of numerical model. < Composite sensitivity pattern of initial SLP and 850hpa temperature fields to the 24h SLP forecast over central Korea (solid box) >

7 CAS/ICSC-8/Doc.X.X, p. 7 < Special observation in ProbeX 2009 > 5. Status Report from Russia 5.1 Russian contribution to the Winter-T-PARC campaign To fulfill, during the winter phase of the THORPEX Pacific Asian Regional Campaign (T- PARC) in January-March 2009 at most 600 additional adaptive radiosonde ascents at 06 and/or 18 UTC on 40 Roshydromet upper-air stations according to the List of participating stations in accordance with the following scheme: Adaptive radiosonde ascents were carried out within six weeks during from ten to fifteen 24-hours intensive observing periods (IOPs). NCEP NOAA/NWS US informed CAO about beginning of the IOPs at least one week before. During each IOP only upper-air stations (about some 20) selected from the List of participating stations, were requested to produce adaptive radiosonde ascents. Selection of upper-air stations and terms 06 and/or 18 UTC for performing adaptive radiosonde ascents will be made by NCEP NOAA/NWS US. (The responsible person is Alexandre Kats, alkats@mail.ru) 5.2 Satellite Data Assimilation A study aimed at studying satellite observation-error covariances and their potential impact in data assimilation was conducted at the Hydrometcentre of Russia. AMSU-A spatial and temporal correlations were objectively estimated. The benefit of accounting for those error correlations in data assimilation was assessed. The main conclusions are: - Horizontal AMSU-A correlations are about as large as background-error correlations.

8 CAS/ICSC-8/Doc.X.X, p. 8 - Inter-channel AMSU-A correlations are also high. - There is significant cross-correlation between background and AMSU-A errors. - There are significant temporal AMSU-A error correlations. - An impact study with simulated data reveals that: (1) Accounting for horizontal and inter-channel observation-error correlations can substantially improve the 3D-Var performance. (2) For a wide range of horizontal scales, the estimated observation-error {\em temporal} correlations do not} lead to a substantial benefit in sequential data assimilation. (Authors V. Gorin and M.Tsyrulnikov, see the 5-th WMO Data Assimilation Symposium's Proceedings) 5.3Adaptive observations in SubArctic East Siberian area addressed to ensemble prediction system (Prepared by Oleg M. Pokrovsky) The aim of this study was to develop and realize a method to determine the optimal locations for new radiosonde (RAOB) sites to be open (reopen) in the data sparse areas and for additional adaptive measurements in many gaps of RAOB network in SubArctic area of Eastern Siberia. The ensemble prediction system (EPS) of a global spectral NWP model was used to determine the most sensitive (to initial data) geographic regions responded to maximum values of the NWP most rapid growing modes in the air temperature and height fields computed by means of the singular value decomposition (SVD) technique implemented in the EPS. The EPS members were investigated in time intervals extending from one to seven days. Non-liner dependences (and its approximations) of the covariance matrices for prediction error fields on the network design (the site numbers and its locations) are discussed. Numerical algorithm to minimize a criterion related to the root mean square (RMS) error field was developed and applied to optimize a configuration of several regional RAOB networks and updating adaptive observations in gap areas. This study is focused to develop an optimal scenario to redesign an existed RAOB by redistribution of stations and network extension in order to minimize the RMS error fields with account to the height, and air temperature fields. Major results were obtained for winter season cases responded to the Arctic cold air outbreaks from high to middle latitudes in Asia, which caused sudden air cool events. Probable impacts of suggested new RAOB and adaptive measurement sites have been evaluated by means of the EPS approach.