Have library access?
IP:3.235.145.252
  • Journals
  • OpenAccess

防災資訊儀表板開發研究

D^2ashboard: A Visual Decision-making Tool for Disaster Prevention

Abstracts


水災的預警須在短時間內分析大量且多維度的水情資料,以臺灣水利決策單位-水利署為例,其目前使用之雨量警戒值淹水預警系統,是利用經驗模式定出全臺各鄉鎮市區之警戒值,當即時雨量超過警戒值即發佈淹水警戒,此系統已應用於近年各颱風豪雨事件之應變操作,結果顯示其在實務運作上能達到預警效果,但其對於觀測雨量、淹水警戒值以及地理資訊等不同來源資料之間並沒有做足適當的連結,導致水災應變人員仍浪費時間在水情資訊之間來回比對與查詢,因此本研究以此系統為基礎,開發一個防災資訊儀表板(D^2ashboard),並針對防災資訊的特性設計互動介面,使用三種資料互動模組:(1)資料提示、(2)資訊刷、(3)動態查詢,解決水情資料無法有效傳遞之問題。本研究於民國102年8月潭美及康芮颱風期間,利用防災資訊儀表板實際進行水災預警與應變作業,結果顯示在兩次事件中,使用防災資訊儀表板能分別幫助應變人員減少25%與61%判讀淹水警戒資訊的時間,因此確實能提升研判警戒資訊的效率。

Parallel abstracts


Decision-making for the issuance of flood warnings is a complex process that requires quick analysis of multi-dimensional hydrological data. The Water Resource Agency in Taiwan has developed a flood alert system that has defined two levels for alerts based on predefined rainfall thresholds. The system tracks rainfall in real time through rain gauge stations in specific locations and subsequently issues flood alerts to the affected counties when the rainfall reaches a predefined threshold. The system has been operational during several typhoons and heavy-rain events and has demonstrated its usefulness in flood prevention. However, in order to judge the flood potential, the analyzers have spent a considerable amount of time exploring critical information due to the lack of connectivity between relative hydrological data in the system. The research led to the development of a visual decision-making tool (called D^2ashboard), which allows users to manipulate related information to enhance their understanding regarding the risk of a floods and to influence decision making. We integrated three interactive functions: (1) data tips, (2) data brushing, and (3) dynamic queries in order to increase its usability. Users are now able to take advantage of its capacity for dynamic exploration and from its intuitive operation. We implemented D^2ashboard to the disaster prevention process during the typhoons TRAMI and KONG-REY (August 2013) for validation. The results show that D^2ashboard decreased the time the experts spend on judging the flood information by 25% and 61% respectively.

References


吳東昇、王藝峰()。
宋爾軒、蔡孟涵、康仕仲、賴進松()。
李明營、周佳、柯亙重、洪志成、翁春雄、許晃雄()。,未出版。
康仕仲、李清勝、何興亞()。,未出版。
康仕仲、李清勝、何興亞()。,未出版。

Read-around