초록 ▼

1. 연구개발 목표
• 중규모 저기압에 동반된 구름무리 (CC-ML) 호우 예측 시스템 개발
• 중규모 기압골에 동반된 구름무리 (CC-MT) 호우의 발생 기구 파악

2. 연구 내용 및 결과
1) 중규모 저기압에 동반된 구름무리 호우 예측 시스템 개발 연구
① 중국과 황해 상에서 발생하는 중규모 저기압의 통계적 조사
• 2006-2014년 6-8월 사이 141 개의 중규모 저기압 사례 진단 및 분석 수행
• 중규모 저기압의 최초 발생 지점을 6 개의 지역 그룹으로 구분하고, 각 그룹 중규

1. 연구개발 목표
• 중규모 저기압에 동반된 구름무리 (CC-ML) 호우 예측 시스템 개발
• 중규모 기압골에 동반된 구름무리 (CC-MT) 호우의 발생 기구 파악

2. 연구 내용 및 결과
1) 중규모 저기압에 동반된 구름무리 호우 예측 시스템 개발 연구
① 중국과 황해 상에서 발생하는 중규모 저기압의 통계적 조사
• 2006-2014년 6-8월 사이 141 개의 중규모 저기압 사례 진단 및 분석 수행
• 중규모 저기압의 최초 발생 지점을 6 개의 지역 그룹으로 구분하고, 각 그룹 중규모 저기압의 발생 빈도, 이동 경로의 특징, 한반도 영향 비율 등을 파악

② CC-ML 호우 예측 시스템 개발
CC-ML 호우 예측 알고리즘을 개발하였으며, 알고리즘에 따라 CC-ML 호우예측시스템을 구축하였음. 예측 시스템은 i)CC-ML 발생 확인 및 한반도 영향 가능성 진단 알고리즘, ii)수치예측 시스템, iii)강수 예보 결정 알고리즘으로 구성됨

③ CC-ML 호우 예측 시스템의 강수예측 능력 평가 및 기상청 현업 강수예보 적용 평가
• 예측 시스템(CFSR-WRF와 기상청 RDAPS-UM)의 CC-ML 강수 예측 능력 평가
• CC-ML 호우 예측 알고리즘의 기상청 현업 강수예보 적용 및 기상청 현업 수치예측을 이용한 CC-ML 호우 예보 방안 제시

2) 중규모 기압골에 동반된 구름무리 호우의 발생기구 파악
① 중규모 기압골과 연관된 강수의 통계적 특성
• 2001-2010년 6-8월 사이 발생한 CC-MT 호우 17개 사례들에 대해 분석을 수행
• CC-MT 호우의 일변화 경향: 새벽 혹은 이른 아침 (5 LST – 8 LST) 사이에 최대치가 나타나고, 2차 최대치는 나타나지 않음
• 한반도 남부지역 및 중부지역에서는 해안가에서 강수량의 최대치가 새벽에 나타나고 동쪽으로 (내륙으로) 갈수록 최대 강수 시각이 늦어짐

② CC-MT 호우 발생의 원인과 과정
• 2 개의 CC-MT 호우사례에 대한 조사를 통해 중규모 기압골과 연관된 구름무리 호우의 발생 원인과 과정 조사

• 발생 원인과 과정의 특징:
- 서태평양 고기압으로부터 연장되어 발달한 기압능이 한반도를 덮고 있을 때 발생
- 동중국해~황해상에서 구름무리 발달 지역까지 하층제트 존재
- 황해에서 발생한 대류계가 북동진하여 서해안으로 접근하면서 중규모 기압골 발달
- 중규모 기압골 발달과 함께 구름무리 발달 등

③ CC-MT 호우 발생에 대한 연구결과의 예보 활용 방안 제시
• CC-MT는 황해 또는 서해안에서 발생하고, 발생과 함께 급격하게 호우가 나타나는 특징을 보임. 본 연구 결과에서 얻어진 발생 과정의 정보를 참조하여, CC-MT 발생의 징후 탐지를 통한 호우 가능성 사전 예보 전략을 제시하였음.

(출처 : 요약서 4p)

Abstract ▼

Research and development have been carried out with the following goals:
∙Development of prediction system for heavy rainfall related to cloud clusters associated with mesoscale lows (CC-MLs)
∙Understanding of genesis mechanism of heavy rainfalls related to the cloud clusters associated with m

Research and development have been carried out with the following goals:
∙Development of prediction system for heavy rainfall related to cloud clusters associated with mesoscale lows (CC-MLs)
∙Understanding of genesis mechanism of heavy rainfalls related to the cloud clusters associated with mesoscale troughs (CC-MTs)

1. Development of prediction system for heavy rainfalls associated with CC-MLs
1) Statistical analysis of mesolows appearing over China and the Yellow Sea
Meso-α-scale lows (MLs) appeared during June-August of 2006-2014 over east Asia including whole China and Mongol area have been identified and their statistical features were examined to reveal their formation areas, tracks with movement speeds, and pattern of tracks toward the Korean peninsula. Major features are as the following:
∙141 MLs were identified during June-August 2006-2014. Mean annual number of MLs was 15.6 and annual number of MLs did not vary greatly.

∙Locations of the first appearance of ML were grouped into 6 areas considering terrain features. Frequency of ML formation for each group: 1) 20 MLs over the area to the southeast of Tibet plateau (R1A), 2) 31 MLs over the area to the northeast of Tibet plateau (R1B), 3) 8 MLs over northern China (R2A), 4) 52 MLs over central and eastern China (R2B), 5) 24 MLs over the Yellow Sea and the East China Sea (R3A), and 6) 6 MLs over southeastern China and Taiwan (R3B).

∙More than half of MLs were found to pass over or near the Korean peninsula except for the MLs formed over R1B(45 %) and R2A(38 %) areas. 54 percent of the MLs formed over central and eastern China (R2B), which was the largest ML source area, reached the areas where MLs could produce rainfall over the Korean peninsula with the average travel time of 34 hours. 71 % of the MLs formed over the Yellow Sea and East China Sea (R3A) moved to the Korean peninsula with average travel time of 19 hours.

2) Prediction system for heavy rainfall associated with CC-ML
An algorithm for the prediction of heavy rainfall associated with CC-ML was developed based on the knowledges obtained in the present statistical study and the previous stage of this project. And a prediction system was proposed to carry out the forecast of heavy rainfall associated with CC-MLs, and was used to predict heavy rainfalls in the events of CC-ML heavy rainfall during 2011-2014. The present algorithm was also used with RDAPS-UM(12km) of KMA for an evaluation of the algorithm in operational forecasts. Details of the present work are as the following:

① Forecast algorithm of heavy rainfalls associated with CC-MLs
The algorithm consists of i) monitoring of ML formation, ii) decision on the CC-ML occurrence, iii) decision on whether CC-ML can affect south Korea, iv) numerical prediction of rainfall associated with the CC-ML, and v) forecast of rainfall associated with the CC-ML. Detailed processes were proposed for each step of the present algorithm.

② Forecast system for heavy rainfalls associated with CC-MLs
Forecast system to carry out the algorithm consists of i) the routine of diagnosing CC-ML formation, ii) numerical weather prediction (NWP) system, iii) the rainfall forecast based on the information of CC-ML from the monitoring and numerical prediction of CC-ML (predicted formation, tracks, rainfalls, etc.). Two forecast systems were used for this study with differing NWP systems: i) WRF model with initial fields from CFSR, ii) KMA’s UM (12km) with RDAPS initialized fields. The second system using the RDAPS-UM is for the application test of the present algorithm to operational forecast.

③ Evaluation of rainfall forecasts
The proposed forecast systems were applied to the forecast of rainfalls in the 23 CC-ML events that occurred in 2011-2014. The evaluation is performed for the threshold values of observed rainfall rates of 25mm/12h and 50mm/12h, based on threat score (TS), bias score (BS), false alarm rate (FAR), etc.

The rainfall forecasts made at the time of the first ML occurrence showed TS greater than 0.3 for the threshold of 25mm/12h in 9 of the 23 cases when CFSR-WRF(12km) is used and 13 cases when RDAPS-UM(12km) prediction is used. RDAPS-UM prediction also showed smaller FAR than CFSR-WRF prediction. Evaluation of forecast for the threshold of 50mm/12h is carried out for 17 cases. The rainfall forecasts made at the time of the first ML occurrence showed TS greater than 0.3 in 7 of the 17 cases when CFSR-WRF(12km) is used, and 3 cases when RDAPS-UM(12km) prediction is used. The higher TS for the CFSR-WRF prediction is partly due to its larger BSs (i.e., BSs were in the range of 1.17 ~1.63 for 4 cases with TS greater than 0.3. On the other hand, BSs for RDAPS-UM prediction were in the range of 0.00~1.02 for the same cases.) The evaluation indicated that CFSR-WRF tended to overpredict the rainfall associated with CC-MLs, while RDAPS-UM(12km) tended to underpredict.

④ Suggestions for the use of the forecast algorithm for rainfall associated with CC-ML
The evaluation of the operational application of the proposed algorithm for heavy rainfall forecast (i.e., application of the algorithm using the prediction of RDAPS-UM) can be summarized as the following:
∙ Performance of rainfall forecast:

∘Threshold of 25mm/12h
Rainfall forecasts at ML formation times showed threat score greater than 0.3 in 13 (57%) of the 23 cases. Performance did not vary significantly with the initial time of NWP.

∘Threshold of 50mm/12h
Rainfall forecasts at ML formation times showed low threat score less than 0.2 for 14 of the 17 cases. However, forecasts made at 12 hour before strong rainfall began in south Korea showed TS greater than 0.3 in 62 % of the 17 cases.
∙ Performance of rainfall forecast varied with the pattern of CC-ML track

∘Rainfall forecasts showed significantly higher TS for nearly straight tracks than those for significantly curved tracks: TS25mm was greater than 0.3 in 8(80 %) of 10 nearly stright track cases, but it was smaller than 0.2 in 7 (70%) of 10 significantly curved tracks.
The above features were considered to improve the rainfall forecast algorithm for use in the operational rainfall forecast as described in the text.

2. Investigation of generation mechanism of heavy rainfall associated with CC-MT
Statistical and case studies were made to understand the heavy rainfall associated with CC-MT: i) 17 CC-MT heavy rainfall cases in 2001-2011 were examined to extract statistical characters of heavy rainfall, and ii) two case studies were carried out to find out the generation processes of heavy rainfall associated with CC-MT. Suggestions were made concerning the use of the findings from these studies for the forecast of heavy rainfall associated with CC-MT.

1) Statistical characteristics of rainfall associated with CC-MT
Maximum of hourly rainfall amount occurred at the dawn or in the early morning (5 - 8 LST) according to the examination of 17 CC-MT cases. Secondary peak was not found. The time of maximum hourly rainfall varied with location: it was found first (at dawn) over the west coast area of middle and southern part of the Korean peninsula and then later in the morning toward the inland area, except for Chungchung and Kyoungbuk areas.

2) Generation processes of heavy rainfall associated with CC-MT
Case studies were made for two CC-MT cases to examine the cause and process of heavy rainfall development in association with CC-MT. Common characteristic features concerning the cause and generation process of heavy rainfall were extracted based on the two cases studies:

∙Heavy rainfall associated with CC-MT occurred over south Korea when pressure ridge extended from the western Pacific subtropical high (WPSH) was found over the Korean peninsula (the ridge did not extend upto the inland China)

∙A low-level jet (LLJ) was present from the East China Sea-Yellow Sea to cloud cluster area (The presence of LLJ was attributed to the strengthening of height gradient between the ridge associated with the WPSH and trough development over eastern China).

∙A mesoscale trough developed near the west coast of the Korean peninsula as convective systems from the southern Yellow Sea approached the west coast
∙Cloud cluster developed over the area of the mesoscale trough at the northern end of the LLJ
∙The mesoscale trough and the associated cloud cluster moved slowly possibly due to the presence of mesoscale pressure ridge at their downstream over inland area

3) Suggestions for the use of findings in the forecast of heavy rainfall associated with CC-MT
The above common features in the generation processes of heavy rainfall can be used for the forecast of heavy rainfall associated with CC-MTs which rapidly develop producing heavy rainfall. Some suggestions have been made for the use of findings from this study in the forecast of heavy rainfall associated with CC-MT.

(출처 : Summary 10p)

목차 Contents

• 표지 ... 1
• 제출문 ... 3
• 보고서 요약서 ... 4
• 요약문 ... 6
• Summary ... 10
• CONTENTS ... 14
• 목차 ... 16
• 표목차 ... 18
• 그림목차 ... 19
• 제 1 장 연구개발과제의 개요 ... 22
• 제 1 절 연구개발의 목적 및 필요성 ... 22
• 1. 구름무리 호우 예측 개선의 필요성 ... 22
• 2. 연구 배경 ... 22
• 제 2 절 연구개발의 목표 및 범위 ... 23
• 1. 연구개발의 목표 ... 23
• 2. 연구개발의 내용 및 범위 ... 23
• 제 2 장 국내외 기술개발 현황 ... 25
• 제 3 장 연구개발수행 내용 및 결과 ... 27
• 제 1 절 중규모 저기압에 동반된 구름무리 호우 예측 시스템 개발 ... 27
• 1. 개요 ... 27
• 2. 중국과 황해 상에서 발생하는 중규모 저기압의 통계적 조사 ... 29
• 3. CC-ML 호우 예측 알고리즘 및 예측 시스템 개발 ... 34
• 4. CC-ML 호우 예측 알고리즘의 기상청 현업 강수 예측 적용 및 평가 ... 39
• 제 2 절 중규모 기압골에 동반된 구름무리 호우의 발생 기구 파악 ... 69
• 1. 개요 ... 69
• 2. 중규모 기압골과 연관된 구름무리 호우 사례 조사 ... 69
• 3. 중규모 기압골과 연관된 구름무리 호우의 발생기구 조사 ... 78
• 4. 중규모 기압골과 연관된 구름무리 호우의 주요 특징과 연구결과의 예보 활용 방안 ... 99
• 제 4 장 목표달성도 및 관련분야에의 기여도 ... 101
• 제 1 절 연구개발 목표달성도 ... 101
• 제 2 절 관련분야에의 기여도 ... 102
• 제 5 장 연구개발결과의 활용계획 ... 104
• 제 6 장 연구과정에서 수집한 해외과학기술정보 ... 106
• 제 7 장 국가과학기술종합정보시스템에 등록한 연구시설∙장비 현황 ... 107
• 제 8 장 연구개발과제의 대표적 연구 실적 ... 108
• 제 9 장 참고문헌 ... 109
• 끝페이지 ... 109