표제지
연구보고서
목차
CONTENTS 7
요약문 21
SUMMARY 25
제1장 서론 29
제1절 연구의 필요성 29
제2절 국내외 관련 연구 31
제3절 연구 목표 및 내용 34
제2장 집중관측과 재해성 기상현상의 기구규명 36
제1절 고분해능 관측자료의 상시 생산 36
1. 국가 악기상 집중관측센터 (NCIO) 운영 36
가. 윈드프로파일러 37
나. 오토존데 39
2. 집중관측 및 상시 관측자료 운영 43
가. KEOP-DB 하드웨어구성 43
나. KEOP-DB 관측자료 구성 43
다. KEOP-DB 관측자료 전송 44
3. 라디오미터의 도입과 운영 54
가. 장비의 특성 56
나. 장비의 운영 및 유지보수 65
다. 현재 운영 현황 69
제2절 집중관측과 재해성 기상현상 구조 분석 71
1. 동계 집중관측 71
가. 기상개황 71
나. 관측자료 효과 분석 79
2. 하계 집중관측 84
가. 집중관측 개요 84
나. 라디오존데 관측 결과 85
다. 장마 특성 분석 88
라. 결론 및 요약 95
3. 선상 집중관측 98
가. 집중관측 개요 98
나. 관측자료 분석 및 효과 분석 99
제3절 해남 관측 자료의 특성 분석 106
1. 라디오미터 자료분석 106
가. 라디오미터와 라디오존데의 관측자료 비교 106
나. 라디오미터와 윈드프로파일러 자료를 이용한 태풍 나리 분석 112
2. 윈드프로파일러와 존데 관측자료의 상호 비교 분석 116
제3장 예측 가능성 분석 118
제1절 고층관측 자료의 주성분 분석을 통한 모형의 예측가능성 118
1. 서론 118
2. 고층관측 자료의 주성분 분석 119
가. 관측자료 119
나. 기온의 주성분 분석 결과 121
3. 수치모델 기준상태 설정에 관한 민감도 연구 123
가. 실험설계 123
나. 민감도 실험 결과 125
4. 요약 및 결론 126
제2절 3DVAR에서 영향반경조절이 분석장과 수치예보에 미치는 영향 128
1. 서론 128
2. 자료 및 실험설계 129
가. 통합 3DVAR를 사용한 분석장 생산 129
나. 실험 설계 130
다. 강수 사례 132
3. 모델예측 결과 및 분석 134
가. 실험결과 134
나. 고층자료에 대한 민감도 135
4. 요약 및 결론 136
제3절 S-밴드를 사용한 다중 도플러 바람장 분석 138
1. 서론 138
2. 다중 도플러 바람장 분석 방법 139
3. 다중 도플러 분석시스템의 검증 142
4. 사례분석 146
5. 요약 및 결론 148
제4장 관측시스템 실험 149
제1절 목표관측 수행을 위한 기반 시스템 연구 149
1. 연구 방법 149
가. 목표관측전략 149
나. 자료동화시스템 152
2. 앙상블 정보를 이용한 수반민감도 기반의 실시간 목표관측전략 162
가. 사례 설명 162
나. 실험 구성 163
다. 결과 166
3. 특이벡터를 이용한 예측 민감 지역 파악 178
가. 사례 설명 178
나. 실험 구성 178
다. 결과 179
4. 자료동화시스템 구축 및 적용 183
가. 사례 설명 183
나. 실험 구성 185
다. 결과 188
5. 결론 및 토의 195
제2절 고층관측 자료의 관측민감도 실험 197
1. 서론 197
2. 관측민감도 실험 197
가. 호우 사례일 : 2007년 7월 4일 197
나. 자료와 수치실험 설계 201
다. 관측자료의 수치민감도 실험 결과 204
3. 결과 207
제5장 요약 및 결론 208
제1절 연구 결과 종합 208
제2절 기대효과 및 활용방안 211
참고문헌 212
부록 219
1. 오토존데 매뉴얼 220
2. 여름철 집중관측자료 236
가. 지상 및 고층일기도 237
나. 집중관측일지 268
3. 한국의 THORPEX/T-PARC 사업 수행에 관한 연구 274
한국의 THORPEX/T-PARC 사업 수행에 관한 연구 275
제출문 276
보고서 요약서 277
요약문 278
SUMMARY 283
목차 288
제1장 서론 292
제1절 연구의 배경 및 목적 292
제2절 연구의 범위 및 방법 293
제2장 KEOP 현황 및 발전방향 295
제1절 KEOP 현황 295
1.1. KEOP 사업개요 295
1.2. KEOP 2007년 연구범위 및 수행방법 305
1.3. KEOP 2007년 연구결과 307
제2절 향후 KEOP 발전 계획 310
제3장 THORPEX/T-PARC 현황 및 계획 313
제1절 THORPEX/T-PARC 배경 313
제2절 THORPEX/T-PARC 현황 322
제3절 THORPEX/T-PARC 관련 국내 인프라 현황 331
제4장 향후 T-PARC 대응 과학적 계획 및 전략 334
제1절 적응관측 현황 및 계획 334
제2절 자료동화 현황 및 계획 337
제3절 OSSEs/OSEs 현황 및 계획 340
제4절 T-PARC 2008 계획 342
제5장 THORPEX/T-PARC 에 대한 한국의 대응 전략 350
제1절 향후 KEOP와 THORPEX/T-PARC 연계전략 350
제2절 향후 GEOSS와 THORPEX/T-PARC 연계전략 353
제6장 결론 및 제언 363
제7장 참고문헌 368
Table 2.1.1. Main devices of windprofiler system. 38
Table 2.1.2. Major specification of windprofiler system. 38
Table 2.1.3. Malfunction report and maintenance procedure of windprofiler in 2006. 39
Table 2.1.4. Major advantage of autosonde. 40
Table 2.1.5. The main purchase parts for autosonde upgrade. 42
Table 2.1.6. Hardware and software information of KEOP-DB system. 43
Table 2.1.7. Monitoring and copy program code of windprofiler (WPR). 46
Table 2.1.8. Monitoring and copy program code optical rain gauge(ORG). 47
Table 2.1.9. Monitoring and copy program code of micro rain radar(MRR). 48
Table 2.1.10. File check program. 49
Table 2.1.11. Scriptsource code in edtfileout key. 51
Table 2.1.12. Scriptsource code in edtfileout key. 52
Table 2.1.13. Digiftp options. 53
Table 2.1.14. The operation situation of the radiometer over the world. 55
Table 2.1 15. The introduction processes of the microwave radiometer. 55
Table 2.1.16. Variables and radio frequency of each model. 56
Table 2.1.17. Specification of microwave radiometer. 58
Table 2.1.18. Calibration method of sensors in microwave radiometer. 63
Table 2.1.19. Detail description of operation program. 65
Table 2.1.20. Check list for the maintenance 68
Table 2.2.1. The information of 2007 winter IOP. 71
Table 2.2.2. The upper observation network during KEOP-2007 in winter and the experimental design for the numerical experiments for the intensive observation data impact. 79
Table 2.2.3. The difference between Loran-C and GPS of radiosonde. 85
Table 2.2.4. The statistics report of radiosonde observation during KEOP-2007. 86
Table 2.2.5. The starting and ending date of Changma in 2007. 89
Table 2.2.6. Experiment designs for the sensitivity test. 102
Table 3.1.1. Configuration of the upper observation station. 120
Table 3.1.2. The numerical designs for the sensitivity experiments 124
Table 3.2.1. Experiment design. 131
Table 3.2.2. Summary of numerical experiments. 132
Table 3.3.1. Configuration of the analysis domain. 143
Table 4.1.1. The definitions of the various stages of the control variable transform given by (4.9) for the unified global/regional WRF-Var system. Indices (i, j, k) refer to grid-point space, index m to vertical mode, and 1, n to global spectral mode. The variables are: u, v: velocity component;... 156
Table 4.1.2. Statistics of (B-O) and (A-O) for each observation type and each observation variable. 190
Table 4.1.3. Statistics (minimum, maximum, mean, and root mean square error) of analysis increment (A-B) for each variable. 190
Table 4.1.4. Statistics of forecast minus observation (F-O) at each forecast time. 191
Table 4.2.1. Numerical experiment configurations. 202
Table 4.2.2. Experiments of sensitivity to observations 203
Table 4.2.3./4.2.4. Rain contingency table. 205
Table 1.2.1. Contents and schedule for research. 294
Table 2.1.1. KEOP and related other IOPs in 2001 and 2002. 301
Table 2.1.2. Same as in Table 2.1.1 except in 2003. 302
Table 2.1.3. Same as in Table 2.1.1 except in 2004. 303
Table 2.1.4. Same as in Table 2.1.1 except in 2005. 304
Table 2.1.5. Same as in Table 2.1.1 except in 2006. 304
Table 2.1.6. Same as in Table 2.1.1 except in 2007. 305
Table 2.1.7. Scope, methods and contents of the KEOP 2007. 306
Table 2.1.8. Research contents and results from the KEOP 2007. 308
Table 2.1.9. Objectives and achievement rate of the KEOP 2007. 309
Table 2.1.10. Self-evaluation for the KEOP 2007. 309
Table 2.2.1. Objectives, contents and scopes of the KEOP 2008. 311
Table 2.2.2. Final goals and contents of the KEOP 2008. 311
Table 3.1.1. Research objectives of sub-programs of THORPEX. 315
Table 3.1.2. Scientific issues and components in T-PARC. 317
Table 3.1.3. Societal impacts and research interests of Asian and North American Regional Committee in T-PARC. 320
Table 3.2.1. Research interests from participating countries in T-PARC 2008. 323
Table 3.2.2. Research purpose, vertical level, instrumentation and platforms for the study of tropical cyclone-midlatitude interface in T-PARC. 325
Table 3.2.3. Same as in Table 3.2.2 except for the study of tropical cyclone/midlatitude impact region. 326
Table 3.2.4. Same as in Table 3.2.2 except for the study of tropical cyclone core. 326
Table 3.2.5. Platforms for observations for three major missions in T-PARC. 328
Table 3.2.6. Facilities, equipment, and other resources for T-PARC. 330
Table 3.3.1. Global prediction system, ensemble forecasting system, regional prediction system at KMA (2008). 332
Table 3.3.2. Research groups related to the THORPEX/T-PARC in Korea. 332
Table 3.3.3. Observation programs currently in operation. 333
Table 4.4.1. Potential research subjects and issues based on data from the KEOP and T-PARC in 2008. 349
Table 5.1.1. A 5-year plan and expected budget for the KEOP and THORPEX. 352
Table 5.2.1. Strategic plan for GEOSS in climate during 2006-2015. 358
Table 5.2.2. Same as in Table 5.2.1 except in meteorology. 359
Table 5.2.3. Same as in Table 5.2.1 except in water resources. 359
Fig. 2.1.1. The placement map of observing equipments at national center for intensive observation of severe weathers (NCIO). 37
Fig. 2.1.2. Autosonde system 41
Fig. 2.1.3. The organization of OPER directory 44
Fig. 2.1.4. Connection flow chart of the equipments at the NCIO. 45
Fig. 2.1.5. EDTFileout values in Outputs/Triggering using DBManager. 50
Fig. 2.1.6. EDTFileout values in Spooling/Messages using DBManager. 52
Fig. 2.1.7. The inner structure and each role of microwave radiometer. 57
Fig. 2.1.8. The vertical distribution of atmospheric absorption spectrum. 60
Fig. 2.1.9. Atmospheric absorption spectra near 22 and 600Hz. 60
Fig. 2.1 10. The receiving and processing data procedure of microwave radiometer. 61
Fig. 2.1.11. The diagram for the data process. 61
Fig. 2.1.12. Neural network process. 62
Fig. 2.1.13. A tripod and chain of microwave radiometer. 63
Fig. 2.1.14. System error of the radiometer at 29 September 2007. 69
Fig. 2.1.15. Pictures of the radome contaminated by dust. 70
Fig. 2.2.1. The mobile observing system of DFM-06 radiosonde. The picture on the left is a receiver and right is a DFM-06 radiosonde. 72
Fig. 2.2.2. The distribution of daily accumulated rainfall during KEOP-2007 in winter. 74
Fig. 2.2.3. Time series of hourly precipitation during KEOP-2007 in winter. 75
Fig. 2.2.4. Surface weather map from 01 Mar. 2007 to 06 Mar. 2007. 76
Fig. 2.2.5. Same as Fig. 2.2.4, but for 500 hPa. 77
Fig. 2.2.6. Time series of average maximum height during KEOP-2007 in winter. 78
Fig. 2.2.7. The skewT/logP diagrams at a) 2330 UTC 02 Mar. 2007, b) 1130 UTC 04 Mar. 2007, and c) 2330 UTC 04 Mar. 2007. 78
Fig. 2.2.8. The horizontal distribution of potential temperature, geopotential height, and wind vector on 850 Ha at 0600 UTC 4 Mar. 2007. 80
Fig. 2.2.9. The horizontal difference fields of geopotential height (blue lines), temperature (red lines), wind speed (green lines), and wind vectors (vectors). The dashed line denotes negative values. 80
Fig. 2.2.10. The vertical profiles of (a) temperature, (b) u-component, and (c) v-component for observation, CNT, and SNT. From the top to the bottom panels are at Munsan, Socho, Pohang, Gwangiu, and Haenam, respectively. 82
Fig. 2.2.11. Thereat scores for 0.1, 1.0, 5.0, and 10.Omm thresholds per 3-hour of (a)CNT and (b) SNT. 83
Fig. 2.2.12. KEOP-2007 observation site 85
Fig. 2.2.13. The time series of average maximum height during KEOP-2007 at each station. 87
Fig. 2.2.14. The distribution of daily accumulated rainfall during KEOP-2007. 90
Fig. 2.2.15. The time series of hourly precipitation at the radiosonde sites during KEOP-2007 91
Fig. 2.2.16. Vertical profile of equivalent potential temperature (shaded) and wind vector (vector), and hourly precipitation at each point. The stations on left panel are located on the southern part of Korea and stations on the right are the middle and northern part of Korea. 93
Fig. 2.2.17. Same as Fig. 2.2.16, but for from 15 Jun. 2007 to 25 Jun. 2007. 96
Fig. 2.2.18. Composite map of equivalent potential temperature (shaded), wind vector (vector), geopotential (m² /s²) for 15-16 Jun. (left), 17-20 Jun. (middle), and 21 Jun. (right). 97
Fig. 2.2.19. Domain of the KLAPS with the navigation route during the observation period and each position of the upper observation site. 98
Fig. 2.2.20. The picture of (a) Ship of "EARDO", and equipments of (b) portable GPS rawinsonde, and (c) mooring. 99
Fig. 2.2.21. Schematic plot for local analysis system with 15 km and 5 km domain configuration. 99
Fig. 2.2.22. Vertical temperature and relative humidity profiles of the observation and KLAPS at each time. 100
Fig. 2.2.23. Same as fig. 2.2.22, but for U-component and V-component. 101
Fig. 2.2.24. (a) The navigation route during the observation period and (b) observation profiles of the temperature and relative humidity at 0300 UTC 26 Aug. 2007. 103
Fig. 2.2.25. The difference profiles between each run (CTL and EXP) and observations (a) temperature, (b) relative humidity, (c) u-component, and (d) v-component. 103
Fig. 2.2.26. The horizontal distribution of the difference between EXT and CTL for temperature (left panels), relative humidity (middle panels), and wind speed (right panels) at (a) 1000 hPa, (b) 850 hPa, (c) 700 hPa, and (d) 500 hPa. 104
Fig. 2.3.1. Comparison of temperature (left) and relative humidity (right) profiles from radiosonde (dashed line) and radiometer (solid line) at 12 UTC (upper) and 21 UTC (down) 21 Aug. 2007. 107
Fig. 2.3.2. Temperature bias of radiometer based on comparison with radiosonde soundings. 107
Fig. 2.3.3. AWS accumulated rainfall on 11-12 Jan. 2008. 109
Fig. 2.3.4. Same as Fig. 2.2.1, but for at 05 UTC (upper) and 12UTC (down) 11 Jan. 2008. 109
Fig. 2.3.5. Water vapor density (left) and liquid water density (right) profiles at 05 UTC (dashed line) and 12 UTC (solid line) 11 Jan. 2008. 110
Fig. 2.3.6. Same as Fig. 2.3.1, but for at 00 UTC 21 Jan. 2008 with wind speed profile from radiosonde. 111
Fig. 2.3.7. Surface weather map at 00 UTC 21 Jan. 2008. 111
Fig. 2.3.8. Relative humidity profiles of radiosonde at Gosan (left) and Heuksando (right) site at 00 UTC 21 Jan. 2008. 111
Fig. 2.3.9. AWS accumulated rainfall on 14-17 Sep. 2007 (left) and the track of typhoon 'Nari'(right). 112
Fig. 2.3.10. The time series of temperature, relative humidity and pressure at surface (left) and of temperature, relative humidity, and water vapor density soundings (right) from 14 to 16 Sep. 2007. 113
Fig. 2.3.11. Integrated water vapor (cross), integrated liquid water (gray line) using the microwave radiometer and the rain rate using the ORG (circle). 114
Fig. 2.3.12. Liquid profile at 10 UTC 14 Sep. 2007 (solid line) and 06 UTC 16 Sep. 2007 (dashed line). 114
Fig. 2.3.13. The time series of CAPE and SREH from 14 to 16 Sep. 2007. 115
Fig. 2.3.14. Temperature profile at 07 UTC and 10 UTC 14 Sep. 2007 and 06 UTC 16 Sep. 2007. 115
Fig. 2.3.15. Relationship between radiosonde and wind-profiler for u (left) and v (right) wind at the NCIO (upper) and Munsan (down). 117
Fig. 3.1.1. Location of upper observation sites during KEOP-2006. 119
Fig. 3.1.2. (a) Vertical distribution of the 1st eigenvectors and (b) associated time series of the 1st loadings of temperature. The shaded gray and white areas mean on/off of Changma front with marking P# during KEOP-2006. (이미지참조) 121
Fig. 3.1.3. Surface weather maps of 4 periods concerned with Changma front on/off. 122
Fig. 3.1.4. (a) Surface weather map, (b) MTSAT enhanced IR image, and (c) 24-hour accumulated rainfall observed by AWS at 1800 UTC June 30 2006. 126
Fig. 3.1.5. ETSs of various six-hourly accumulated precipitation for each run. 126
Fig. 3.2.1. Schematic diagram of analysis and forecasting system with KWRF and U3VR. 130
Fig. 3.2.2. The domain f3r the experiment. 131
Fig. 3.2.3. Location of radiosonde sites during KEOP-2007. 132
Fig. 3.2.4. Three-hour surface weather maps (upper panels), MTSAT enhanced IR imageries (middle panels), and radar imageries (lower panels) from 0000 to 0900 UTC 4 July 2006. 133
Fig. 3.2.5. Hourly precipitation (mm/hr) during experiment period. 133
Fig. 3.2.6. ETS distribution of 6-hour accumulated rainfall at 6-hour forecast for EXPI (left), EXP2 (center), EXP3 (right) [Thresholds are 10, 15, 20, 25mm (upper panels) and 5, 10, 20, 40mm (lower panels)]. 135
Fig. 3.2.7. ETS distribution of 6-hour accumulated rainfall at 6-hour forecast (upper panels, thresholds are 5, 10, 20, 40mm) and deviation between experiments (EXPI, EXP2, EXP3) and experiment without variational analysis in 850 hPa temperature average for domain (lower panels). 136
Fig. 3.3.1. The multiple doppler radar wind analysis domain. The shading area is covered by more than two S-band radars. The big dots are the positions of S-band radars and the small dots are the positions of C-band radars. The hollowed dots are the positions of the wind profilers. The thin circles are the ranges of the S-band radars, and the thick circles are the... 142
Fig. 3.3.2 The scatter diagrams between observed and retrieved wind components. The first column is the scatter diagram for the u wind component, the second column is for the v wind component, the third column is for the wind speed, and the fourth column is for the wind direction. The first row of this figure is for the Munsan wind profiler observation, the second row is for the Kangreung... 144
Fig. 3.3.3. The correlation coefficients between observed and retrieved wind components... 145
Fig. 3.3.4. The water vapor image of the MTSAT and 500 hPa map at 00 UTC on July 29, 2007. The black area in this figure is dry region and the white area is moist region. 146
Fig. 3.3.5. The distribution of surface temperature (left), one hour accumulated precipitation (middle), and sea level pressure (right) overlapped with 10 m wind vector at 1400 KST July 29, 2007. 147
Fig. 3.3.6. A sample of the multi-doppler radar analysis. The analysis time of upper panels is 1400 KST July 29, 2007 and that of are lower panels are 1430 KST July 29, 2007. The left panels are the retrieved wind vector and reflectivity factor at 2000 m level. The middle panels are the cross section on the line AB in the upper left... 147
Fig. 4.1.1. Sketch showing the relationship between dataset (circles) and algorithms (rectangles) of the ARW system. 153
Fig. 4.1.2. Sketch of the role of Stage 0 converters in transforming model-specific data (e.g., ARW, KMA global model, etc.) to standard perturbation fields and relevant metadata (e.g., latitude, height, land/sea, etc.). 160
Fig. 4.1.3. (a) Best track of Typhoon Ewinia and (b) important times for adaptive observations. 165
Fig. 4.1.4. Best track of Typhoon Ewinia and model forecast track of control experiment 167
Fig. 4.1.5. Mean sea level pressure (sloid line) and 850 hPa temperature (dashed line) for control experiment at (a) t=Oh, (b) t=6h, (c) t=12h, and (d) t=24h. The box in (d) indicates the region where the response function is defined. 168
Fig. 4.1.6. Best track, ensemble tracks, and model forecast for real-time experiment of typhoon Ewinia. Response-m track and response-p tracks indicate ensemble members for which the response functions are defined. 169
Fig. 4.1.7. Mean sea level pressure (sloid line) and 850 hPa temperature (dashed line) for real-time experiment at (a) t=Oh, (b) t=6h, (c) t=12h, and (d) t=24h. The box in (d) indicates the region where the response function is defined. 170
Fig. 4.1.8. Vertical distributions of the adjoint sensitivities at t=Oh for control experiment:... 171
Fig. 4.1.9. Vertically integrated adjoint sensitivity distributions and mean sea level pressure (solid) at (a) t=Oh, (b) t=6h, (c) t=12h for control expreiment. 172
Fig. 4.1.10. Adjoint sensitivity distributions at 500 hPa (shade) and mean sea level pressure (solid) at t=Oh for control experiment. 172
Fig. 4.1.11. Vertical distributions of the adjoint sensitivities at t=Oh for real-time experiment using response-m ensemble member:... 174
Fig. 4.1.12. Vertically integrated adjoint sensitivity distributions and mean sea level pressure (solid) at (a) t=Oh, (b) t=6h, (c) t=12h, and (d) t=24h for real-time experiment using response-m ensemble member. 175
Fig. 4.1.13. Adjoint sensitivity distributions at 500 hPa (shade) and mean sea level pressure (solid) at (a) t=Oh for real-time experiment using response-m ensemble member. 175
Fig. 4.1.14. Vertical distributions of the adjoint sensitivities at t=Oh for real-time expriment using response-p ensemble member:... 176
Fig. 4.1.15. Vertically integrated adjoint sensitivity distributions and mean sea level pressure (solid) at (a) t=Oh, (b) t=6h, (c) t=12h, and (d) t=24h for real-time experiment using response-p ensemble member. 177
Fig. 4.1.16. Adjoint sensitivity distributions at 500 hPa (shade) and mean sea level pressure (solid) at t=Oh for real-time experiment using response-p ensemble member. 177
Fig. 4.1.17. Best track for typhoon Rusa (courtesy of Korea meteorological administration) 179
Fig. 4.1.18. Mean sea level pressure at (a) 12 UTC 30 August 2002 and (b) 12 UTC 31 August 2002. The box in (b) indicates the region where the projection operator is defined. 180
Fig. 4.1.19. Singular values of singular vectors for 24 hours run. 180
Fig. 4.1.20. Horizontal distributions of singular vectors at 500 and 850 hPa:... 181
Fig. 4.1.21. Vertical distributions of singular vectors:... 181
Fig. 4.1.22. Horizontal distributions of evolved singular vectors at 500 and 850 hPa:... 182
Fig. 4.1.23. Vertical distributions of evolved singular vectors:... 183
Fig. 4.1.24. Vertically integrated energy-weighted singular vector distributions:... 183
Fig. 4.1.25. KMA surface weather map (a) at 00 UTC 1 and (b) at 00 UTC 2 July 2007. 184
Fig. 4.1.26. KMA weather map (a) on 500 hPa and (b) on 200 hPa at 12 UTC 1 July 2007. 184
Fig. 4.1.27. 24 hour accumulated precipitation (a), (b) at 00 UTC 1 July 2007 and (c), (d) at 00 UTC 2 July 2007.... 185
Fig. 4.1.28. Distributions of selected observations:... 187
Fig. 4.1.29. schematic configuration of experiments. 188
Fig. 4.1.30. ▽J according to the minimization steps. 189
Fig. 4.1.31. KMA radar images from (a) 00 UTC 1 2007 to (i) 00 UTC 2 July 2007, with 3 hour interval. 192
Fig. 4.1.32. Sea level pressure (solid line, hPa) and 6 hour accumulated precipitation (shaded, mm) for CYCL (left panels), COLD (middle panels), and NDOWN (right panels) experiments with 6 hour interval (from (a) 00 UTC 1 July 2007 to (1) 18 UTC 3 July 2007). 193
Fig. 4.1.33. 24 hour accumulated precipitation for (a), (b) CYCL, (c), (d) COLD, and (e), (f) NDOWN.... 194
Fig. 4.2.1. (a) Surface weather map at 09 KST 4 July 2007, (b) observed 12-h accumulated rainfall amount (00 KST~12 KST) and time series at Mungyung and Anyang, (c) enhanced IR, and (d) reflectivity at 06 KST. 198
Fig. 4.2.2. Enhanced IR images (left panels), radar reflectivity (middle panels), and observed AWS rainfall amount (right panels). 199
Fig. 4.2.3. Simulated rainfall amounts from NWP models during the event case. 200
Fig. 4.2.4. Location of KEOP-2007 radiosonde sites. 201
Fig. 4.2.5. A nested grid system included two model domains of 10 and 3.3 km horizontal grid spacing. 202
Fig. 4.2.6. Experimental design. 203
Fig. 4.2.7. Simulated 12-h accumulated rainfall amounts for each experiment. 205
Fig. 4.2.8. Forecast skill scores 12-h rainfall amounts for each thresholds. 206
Fig. 2.1.1. Location of radiosonde sites for the KEOP 2007. 306
Fig. 2.2.1. Comparison of radiosonde sites between the KEOP 2007 and KEOP 2008. 310
Fig. 2.2.2. Field experiments as components of the KEOP in the future. 312
Fig. 3.1.1. An overview of the T-PARC experimental design, including facilities explicitly deployed for the experiment and collaboration with other field campaigns taking place in 2008. 318
Fig. 3.1.2. Traditional vs. interactive forecast systems. 319
Fig. 3.1.3. Targeting observation for typhoon. 321
Fig. 3.1.4. Strategy of targeting observation near typhoon. 322
Fig. 3.2.1. Example of DOTSTAR dropwindsonde data imbedded in the typhoon bogusing. From Harr (2006). 328
Fig. 4.1.1. Schematic diagram of adaptive (targeted) observation. 334
Fig. 4.1.2. Comparison of targeted observation strategies using DOTSTAR dropwindsonde data. 335
Fig. 4.1.3. Same as in Fig. 4.1.2 for typhoon Shanshan (2006). 335
Fig. 4.1.4. Same as in Fig. 4.1.2 for typhoon Ewiniar (2006). 336
Fig. 4.1.5. Planned developments at the Japan meteorological agency prior to T-PARC in 2008. 337
Fig. 4.2.1. Schematic diagram of the 4D-Var. 338
Fig. 4.2.2. Schematic diagram of the incremental 4D-Var. 338
Fig. 4.2.3. Effect of dropsonde assimilation in typhoon forecasts. 339
Fig. 4.3.1. Schematic diagram of OSSE at NCEP.... 340
Fig. 4.3.2. Schematic diagram of OSSEs and OSEs using the DOTSTAR dropwindsonde data. 341
Fig. 4.3.3. Selection of dropwindsonde data (left) for the OSSE and typhoon track forecasts (right). 341
Fig. 4.3.4. Application of the OSE method for typhoon forecasts. 342
Fig. 4.4.1. T-PARC/TCS-08 consolidated calendar of milestones and key events. 343
Fig. 4.4.2. Global operations for the T-PARC/TCS-08. 344
Fig. 4.4.3. Components of the T-PARC/TCS-08. 344
Fig. 4.4.4. Predictability related to introduction of a tropical cyclone in the wave packet. From Harr (2007). 345
Fig. 4.4.5. Sample mission scenarios for the NRL P-3B mission. 346
Fig. 4.4.6. Sample mission scenarios for combined operation of the NRL P-3B and FALCON mission. 346
Fig. 4.4.7. Three major missions for the T-PARC 2008. 348
Fig. 4.4.8. Falcon plans for the T-PARC 2008. 349
Fig. 5.0.1. Research foci in Asian THORPEX. 350
Fig. 5.1.1. Connection between the KEOP and THORPEX. 352
Fig. 5.2.1. Strategy in GEOSS. 354
Fig. 5.2.2. Schematic of the GR0 information system. 355
Fig. 5.2.3. Connection between GEOSS and THORPEX. 362