1. Introduction

     The scientific meetings in Taipei, Chung-Li and Hsinchu, associated with the NCAR field trip to Japan and Taiwan, were very hurried owing to our weekend arrival in Taipei. This schedule allowed only two business days for meetings and visits. However, the meetings and tours of various facilities provided a very comprehensive introduction to the work being done in Taiwan in the fields of meteorological/atmospheric research, weather forecasting, atmospheric remote sensing, especially using GPS, and the upcoming COSMIC (ROCSAT-3 as it is known in Taiwan) GPS radio occultation mission. In addition, we learned about the remote sensing infrastructure (at the Remote Sensing Research Institute), as well as the computer facilities (at the National Center for High-Performance Computing), being developed to support the technical and scientific development of Taiwan.

     The following is a list of the institutes and facilities we visited from 12 to 13 July 2004: the Central Weather Bureau (CWB), including the Satellite Meteorology Center and Forecasting Division, the National Science Council (NSC), the National Space Program Office (NSPO), including the Test and Integration Facility and Multiple-Mission Control Center, the National Center for High-Performance Computing (NCHC), the National Central University (NCU), including the VHF radar and the Doppler weather radar facilities, and the Remote Sensing Research Institute (RSRI). Aside from the guided tours of the various facilities, the afternoon of 13 July was devoted to talks on the research being conducted by the faculty and students in the Department of Atmospheric Sciences of NCU. The U.S. students also gave presentations on their research.

2. The Taiwan Central Weather Bureau (CWB):

     Our first scientific visit in Taiwan was to the headquarters of the Central Weather Bureau (CWB) in Taipei. The CWB is Taiwan's counterpart institution to the National Weather Service (NWS) in the U.S.; its functions include developing meteorological technology (instrumentation), conducting research in meteorology, seismology and astronomy, and administering observations and forecasts of the weather and climate, as well as earthquakes. Organizationally, the CWB is under the jurisdiction of the Ministry of Transportation and Communications, and is headed by a Director-General with assistance from two Deputy Director-Generals. We were told that the Director-General of the CWB has a very challenging job; the President's office sometimes calls him and wants to know the exact trajectory of a typhoon, which may have strong impact on Taiwan.

     We arrived at the CWB around 9:15 Monday morning (12 July), and were received by Mr. Jiang-Ling Hsin, the Director-General of CWB, and his associates. Dr. Cheryl Terng, the coordinator for our CWB visit accompanied us during the entire tour. Director Hsin gave a brief welcoming speech (in Chinese, interpreted into English by Dr. Bill Kuo), which was followed by an overview of the CWB by Dr. Chang. We learned that the CWB is staffed with more that 600 people, most of whom are well-trained technical staff. Since Taiwan has high-impact weather (Mei-Yu fronts, typhoons, etc.) throughout the entire year, as well as very frequent seismic activities, the CWB has a formidable task.

     Our main interests at the CWB are the Meteorological Satellite Center (MSC) and the Weather Forecasting Center (WFC). A total of four presentations were given at CWB, with the first two giving us a general introduction to the operations and research work at these two centers; the last two talks focused more on technical aspects related to COSMIC data management and data assimilation efforts.

2.1 The Future of the Satellite Meteorology Center (MSC), by Dr. Paul Chiou: (CWB):

     In his talk, Dr. Paul Chiou introduced the satellite constellation, from which MSC has been receiving observations since 1981. Currently, MSC is receiving data from GOES-9, FY-2B and TERRA. After 2005, MTSAT-1A observations will also be received. The satellite data are used to monitor the atmosphere, ocean and land. The MSC routinely provides 131 products to various governmental and commercial organizations.

     The MSC is also involved in work related to "Very Short-Range Forecasts" (VSRF) and has an 8-year plan (started in 2001) to improve VSRF in Taiwan. This effort includes 0-2-hour severe weather forecasting using WDSS-QPESUMS (Warning Decision Support System-Quantitative Precipitation Estimation and Segregation Using Multiple Sensors), a weather decision support system based on quantitative precipitation estimation from Doppler radar observations, and 3-12-hour rainfall forecasting using a mesoscale model based system, LAPS-MM5. With coverage from four Doppler radar sites over Taiwan, WDSS-QPESUMS gives instantaneous estimate and extrapolation of the precipitation field for the entire island, and provides storm identification and tracking. LAPS-MM5 has a cloud analysis scheme and performs data fusion and then short-term forecasts. The VSRF products, combined with the use of geographical information systems (GIS), are published on the CWB website.

2.2 Overview of Taiwan's Weather Forecasting Center (WFC), by Dr. T.-J. Yeh:

     This talk introduced the major types of weather systems found around Taiwan, their characteristics, and Taiwan's topography and its impact on weather observations and forecasting. As an island of only 200x400 km, even a small forecast error of a typhoon's trajectory or the location of a Mei-yu front, along with their interaction with the topography, can have a significant impact on a weather forecast. The WFC routinely issues 36-hour weather forecasts four times daily, 7-day inland outlooks once daily, 1-month and 3-month long-term major system and trend outlooks, as well as 3-day weather forecasts for the fishing industry. Severe weather forecasts are issued on an as-needed basis. Sea warnings are issued every three hours when a typhoon is expected to affect the seawaters within 100 km of Taiwan.

     The current guidance for weather forecasting at WFC comes from NWP models, Model Output Statistics (MOS), analog models, and the Weather Integration and Nowcasting System (WINS). The WFC is also actively involved in research in quantitative precipitation forecasting using radar, satellite and numerical models. Currently, four numerical weather prediction models are routinely run at the WFC: global, regional, and mesoscale models, and a typhoon-tracking model.

2.3 The Taiwan Analysis Center for COSMIC (TACC), by Mr. Mark Cheng:

     The plan for TACC is to build a mirror computing center for the COSMIC Data Analysis and Archive Center (CDAAC) in Boulder, Colorado. The TACC will be responsible for incoming data translation and quality checking, data conversion and product validation, as well as data distribution and archiving. The TACC will generate results (products) and distribute its data and products to a broad range of scientific and operational users.

2.4 Assimilation of GPS Radio Occultation Data, by Dr. Jen-Cheng Chang, Chinese Culture University and Dr. Ming-Dean Cheng, Director of the Research and Development (R&D) Center, CWB:

     This talk touched upon the on-going research at the CWB on the use of GPS data. Experiments are performed to explore the impact of assimilating radio occultation (RO) data into global and regional numerical models. The bending angle data, as well as the refractivity data, are assimilated into the global model. Assimilation of refractivity alone is tested for regional analyses. Many of these concerns were the subject of discussion at the GPS Colloquium in Boulder prior to the field trip.

     For the regional analysis, the Weather Research and Forecast (WRF) model and a three-dimensional variational assimilation (3DVAR) scheme are used. For the global model analysis, a 2D GPS ray-tracking scheme is incorporated into the data assimilation system. Experiments are conducted using 3DVAR and 4DVAR schemes with CHAMP data. This work is on-going and the future plan includes writing a local refractivity operator, developing data thinning methods, and speeding up the computations.

3. National Science Council (NSC):

     At 11:15AM, we left the CWB and drove to the National Science Council (NSC) building, also located in downtown Taipei. A meeting was arranged for us with Dr. Wei-Chung Wang, Director-General of the Department of International Programs (NSC/DIP) of NSC. Dr. Wang hosted a luncheon for us, during which we had discussions about existing and potential future opportunities for collaborative research between the U.S. and Taiwan. During the lunch, an NSC staff member gave a talk on the general state of research and development in Taiwan, then Dr. Wang talked about the commitment and efforts of the NSC/DIP to establish international scientific cooperation.

     NSC is a cabinet-level organization, equivalent to our National Science Foundation, that primarily manages research and development affairs in science and technology. Major missions of the NSC include promoting national science and technology development, supporting academic research, and developing science-based industrial parks. The Department of International Programs (DIP) serves as the liaison for international scientific cooperation within the NSC. The NSC has formally established collaborative ties with many countries including the U.S. An agreement signed by NSC and U.S. National Science Foundation (NSF) has become an umbrella agreement for all scientific and technological cooperative agreements between the U.S. and the Republic of China. COSMIC is one such on-going cooperative program between NSF and NSC.

     One effort of NSC/DIP is the implementation of Visiting Programs and International Seminars. An example is the Summer Institute in Taiwan for French and U.S. graduate students. Beginning in the year 2000, a program was established to support up to twenty U.S. graduate students to visit Taiwan for eight weeks during the summer. The NSC provides logistical support during the students' stay in Taiwan, while the NSF provides international airfare and a living allowance for the U.S. students. Possible Taiwanese host institutions include national laboratories, governmental scientific and technical units, and research institutes, universities and private sector research centers.

4. National Space Program Office:

     At the National Space Program Office (NSPO), we were shown the Multiple-Mission Control Center (MMC) and the Test and Integration Facility. The MMC is the facility from which the ROCSAT-3 (know in the west as COSMIC) mission will be controlled. The Taiwanese government seems to consider the development of their space program as a vital element of the overall scientific and technical progress of the country. The MMC is currently controlling the ROCSAT-1 and ROCSAT-2 missions; although I believe that ROCSAT-1 had nearly (or just recently) completed its mission at the time of our visit. The ROCSAT-2 satellite was the first spacecraft to detect "sprites" from orbit, Sprites are usually red, luminous emissions appearing above active thunderstorms and associated with cloud-to-ground lightning.

     At the Test and Integration Facility, we had a chance to see the assembly of two of the six COSMIC spacecraft. The facility included sections to test for thermal and acoustic stress, as well as instrumentation to verify the working of all the spacecraft components after integration. The six COSMIC satellites will fly in three orbital planes (two satellites per plane). After launch, the pairs of spacecraft, each carrying only 6.5 kilograms of fuel, will attain different orbital altitudes. This maneuver will allow the orbital planes, over the course of about 12 months, to precess the desired amount relative to one another; the satellites are then raised to the same circular orbital altitude. Since many of the students and junior scientists work with GPS radio occultation data currently, and will likely use COSMIC data in the future, this was a most interesting tour.

5. National Center for High-Performance Computing (NCHC):

     Only forty minutes were allotted to visit the NCHC. However, we were introduced to two major Taiwanese initiatives. The Taiwan Advanced Research and Education Network (TWAREN) is Taiwan's bid to become the Internet gateway to Southeast Asia. The infrastructure of TWAREN is currently under construction. The TWAREN is expected to make possible a distributed computing and resource sharing grid, known as the Knowledge Innovation National Grid (KING). It is hoped that these two projects, working in conjunction, will promote national and international collaborative research, while reducing the cost of such research.

6. Space and Remote Sensing Institute (SRSI):

     Only one hour was allocated for our visit to the Space and Remote Sensing Institute, but we were treated to some interesting visual data representations. Upon arrival at SRSI, we took a 3-dimensional (3D) virtual "fly-over" tour of Taiwan. We wore polarizing glasses image (see photo below) to give the impression of a 3D. Starting from a view of SRSI, we zoomed in on the university campus; we then zoomed out and took a tour clockwise around Taiwan. The most spectacular part of the demonstration was the image detail. Another lasting impression is that the island of Taiwan is so green; even though Taiwan is densely populated, so much of the topography is still covered in lush, beautiful forest.

     Students wearing polarizing glasses to view a 3D virtual "fly-over" of the island of Taiwan, presented at the Remote Sensing Research Institute. (Photo courtesy of Dione Rossiter.) The scientists leading the demonstration then described SPOT-2, a French-made earth observing, remote sensing satellite, launched on January 22, 1990. SPOT-2 not only observes the Earth's surface immediately below the satellite, but also along a slantwise path relative to the satellite, by changing the scan direction of the sensors. SPOT-2 has a resolution of about 10 meters, but the speaker mentioned satellites having a resolution of up to 16cm; these images, however, are not available to the public. We also saw real-time incoming images of the Peoples' Republic of China, South Korea and Japan from the SPOT-2 satellite. Images are generally ordered several weeks prior to a pass over of the region of interest, and the appropriate instructions are sent to the spacecraft. The speaker did indicate that for a "time-critical" event, instructions could be sent to the spacecraft with only several hours' notice.

7. National Central University (NCU):

     Although the main purpose of our visit to NCU was to learn about the research being conducted there, and to present our own research, we had a little time to visit the facilities used to train students in taking meteorological measurements (traditional and more "state of the art" instruments). Professors Y.-H. Liu and P.-L. Lin described these facilities, including the Chung-Li 52 MHz. VHF radar and Doppler weather radar. The VHF radar is similar in configuration to the Middle and Upper Atmospheric Radar (MU), which we visited in Japan. One of Dr. Lin's students later talked on research to assimilate these radar data into models.

     The president of NCU, Dr. Chuan Sheng Liu, greeted each of us during a luncheon arranged for our visit. Many of us were impressed that the president of a university would take the time to greet visitors personally; most of us have never met the president of our own universities. Dr. Liu also opened the "Mini-Workshop for COSMIC Outreach Program", where NCU faculty, and Taiwanese and U.S. students presented talks on their research. Few students presented talks directly related to the use of GPS data, and those who did were not very far along in their research. Still, this was a good opportunity to examine the new frontiers in data assimilation and numerical weather modeling.

     Dr. T.-C. Chen discussed how forecasting for the subsequent four hours, so-called "Nowcasting", is showing improvements using data from Taiwan's six Doppler radar sites. Dr. Chen also investigates the assimilation of radar data into models. Through a simulated radar observation experiment, he has shown that a one-dimensional radar observation of wind can be effective in improving a three-dimensional wind forecast, even in the vertical.

     Dr. Y.-C. Liou presented a method for determining two-dimensional (2D) wind velocity in typhoons using a single Doppler radar. By assuming that winds are mostly circular around the typhoon's eye, it is possible to recover the general 2D-wind structure of the storm with the loss of some detail. Dr. Mei Xu spoke next about her work at NCAR's Research Applications Program (RAP); she is attempting to add cloud scale information into a larger grid size mesoscale model. Her program assimilates radar data into the MM5 model to determine which assimilation methods are effective.

     Dr. P.-L. Lin discussed "Dropsonde Observations for Typhoon Surveillance near the Taiwan Region" (DOTSTAR), which uses instrumentation dropped from airplanes into typhoons to improve forecasts. Use of data from the program has demonstrated improvement to the U.S. Navy's NOGAPS model. Because of these results, the DOTSTAR program may be expanded in the near future.

     Dr. Y.-H. Chu detailed the current missions of Taiwan's space program, focusing on ROCSAT-2's ability to detect "sprites" and "jets" in the upper atmosphere. In addition to the satellite missions, the Taiwanese space program has launched rocket-sondes over the last 15 years to detect changes in the climate of the stratosphere and mesosphere.

     Graduate student S.-H. Lin gave a short discussion about his work in data assimilation of Doppler radar winds. He assimilates only the radial wind, along the line of sight from the radar, which requires many fewer assumptions than methods discussed above.

     Kelly Amundsen, from Ohio State University, presented her research on storm formation in Eastern Antarctica, that portion of Antarctica that lies in the Eastern Hemisphere. Regarded by some as a region of little weather activity, Amundsen and her colleagues believe that MODIS data confirm mesoscale cyclogenesis in the region.

     Erick Adame, of the State University of New York at Oneonta, plans to use radio occultation data to examine the change in tropopause height over Antarctica. Data from GPS/MET, CHAMP, SAC-C, and COSMIC satellites establish a climate record, which can be used to investigate climate change in the southern continent.

     Following the Antarctic presentations, graduate and undergraduate students from NCU, without extensive presentations, briefly discussed their research interests. Susan Wang presented a case study of a typhoon where data assimilation of certain parameters would have significantly improved the quality of the forecast. Kuo-Feng "Stephen" Wang discussed large-scale ionospheric disturbances, which can disrupt communications. Chih-Ying Chen is working on modeling typhoons using the Weather Research and Forecasting (WRF) model. Vicky Fang, a recent graduate at NCU, discussed historical statistics which showed that periods of high solar sunspot activity strongly correlate to natural disasters on Earth. Understanding this phenomenon may deepen our knowledge of climate and help us plan for years when natural disasters may become more common.

     Brent Laabs, from the University of California at Davis, followed this discussion with a presentation on his plans to use GPS radio occultation data from COSMIC to improve weather forecasts in California and the Eastern Pacific. He will use the WRF model's 3D variational analysis (3DVAR) to assimilate the COSMIC data. The COSMIC data set will provide greatly needed observations over the oceans.

     Mike Duda, of the University of Colorado at Boulder, discussed the Antarctic Mesoscale Prediction System (AMPS), developed at the Mesoscale and Microscale Meteorological Division of NCAR. He has primarily worked on the technical aspects of the system, such as running the MM5 software on different platforms, networking computers, and determining the most effective method to present data.

     After the last American student finished, Cheng-Yung "Kevin" Huang addressed the use of ground-based GPS observations in forecasting typhoon rainfall. He compared MM5 forecast precipitation with actual rainfall rates, which he correlated with the GPS PWV measurements. The retrieved PWV appears to be a good indicator of actual rainfall rates. He suggests that GPS observations be used by the Central Weather Bureau (CWB) to assist in "Nowcasting".

     Carlo Wang concluded our mini-workshop with a discussion of sources of air pollution in Taiwan. By tracking the trajectories of air parcels on high pollution days, Wang has discovered three regions, which contribute to much of Taiwan's pollution: large sandstorms coming from the Gobi Desert, industrial pollution from the Peoples' Republic of China, and biomass burning to clear land for agriculture in Southeast Asia. Each type of pollution has its own signature and source region. Knowing the path of weather events from these source regions may allow the CWB to better predict high pollution events.

8. Scientific Experience; Prospects for Future Collaboration:

     The U.S. participants of the field trip had varied expectations on the possibility of future collaborations, based on their experiences during the scientific meetings. Certainly this can partially be explained by the individual student's level of knowledge and ability to understand the material presented. It is also governed by the level of fatigue, for example due to jet lag or the intense schedule. In this section we give representative, but not exhaustive, responses of the students; the material is edited for clarity and to reduce redundancy, but hopefully retains the intent of the students.

     This field trip, in conjunction with the NCAR colloquium on GPS atmospheric remote sensing, presented us with the opportunity to see COSMIC in the making. We met representatives of the organizations funding COSMIC (or ROCSAT-3 as it is known in Taiwan), and met designers of the instruments in Boulder. We viewed two of the satellites (numbered 2 and 3) being assembled in the integration and test facility at NSPO, as well as the mission control room for these satellites.

     At UCAR/NCAR we discussed the characteristics of the GPS data that COSMIC will produce, and how that data will be incorporated into models for numerical weather prediction. At NCU, we discussed the improvements accomplished in weather prediction models using space-based GPS data from CHAMP and SAC-C. In Japan and Taiwan, we visited the national weather agencies that will be using COSMIC data in their weather forecasts for the general public. For those of us who are just beginning our careers as graduate students, the field trip was an excellent opportunity to see where to begin our research, and obtain an understanding of what our roles might be in the field of GPS atmospheric research.

     One of our group (Brent Laabs) was very sure of the importance of the field trip and the likelihood of his further collaboration. We quote his report verbatim: "I feel as though I may collaborate with some of the Taiwanese researchers at NCU for some of my projects in the next few years. My advising professor, Dr. Shu-Hua Chen, has already established relationships with many of the professors there, and it seemed as though everyone there considered her a friend and colleague. So it seems that there is good reason to suspect that I will be interacting with those students again soon. It was very helpful to meet all of these scientists that I may be working with in the future. I enjoyed my stay in Taiwan so much that I may even consider studying there for a year."

9. Cultural Experience; Prospects for Future Collaboration:

     The following are some impressions on the cultural aspects of our visit to Taiwan. Including Taiwanese students in our program was very helpful, and improved our cultural experiences greatly. On the most fundamental level, it meant that at least half of our party spoke the local languages, either Mandarin, Cantonese or Taiwanese. With the help of our student hosts from NCU, we were able to explore a great deal of Taipei, Danshui and Hsinchu. From haggling at the Jade market to showing us the trendiest nightclubs, the NCU students showed us a slice of life that we could have never discovered alone in so short a time.

     They were not only excellent translators and tour guides, but as students we had much in common. It is clear that we share so many of the same desires, goals and hurdles, such as: academic careers, theses to write, and research to do. They were easy to relate to, despite the obvious cultural differences.

10. Concluding Remarks:

     This field trip was a fantastic experience, especially for the younger students. Such experiences will encourage them in their current studies and may lead to ongoing international cooperation in their future careers. We can hope that, as a nation, the U.S. will continue to promote international cooperation. We as Americans can sometimes be provincial, and therefore ignorant of what is going on in the world. During the GPS colloquium in Boulder, Bill Kuo mentioned the abbreviation "NIH", meaning "not invented here". Although this phrase was meant to apply to institutions, Americans tend to apply this idea to the world. This field trip educated the students to the fact that innovative research and technical progress are being made in other parts of the world. Indeed, some of the most interesting projects using GPS for atmospheric research are being developed in Taiwan and Japan. Examples of these projects are the COSMIC satellite project in Taiwan, and GEONET and "downward looking" ground- and aircraft-based GPS stations in Japan. There is no doubt that this field trip was beneficial to both groups, the Taiwanese as well as the U.S. contingents.

     On behalf of our entire group, we would like to thank the people at UCAR, especially Scott Briggs, Bill Kuo and Kim Prinzi-Kimbro, for encouraging us to participate in this field trip. We would also like to thank the Department of International Programs of the National Science Council of Taiwan, and its Director General Dr. Wei-Chung Wang, for sponsoring this field trip.