14.5.5
SDG 14.5.5
- Urban Flood Warning System都市溢淹示警系統
In view of the increasing severity of urban flooding disasters in recent years, urban disaster prevention and warning work is one of the important tasks of the Construction and Construction Department. GIS center assists in the establishment of the ‘urban flood warning system’. Based on the design standards of stormwater sewers in various regions, the rainfall warning value of the sewer system is set, and the urban flood warning of 322 townships (towns, cities, districts) in Taiwan is provided. Information.
The main functions of this system include current status warning, future warning, report, setting, reference material, CEOC area, advanced management, etc.; use map application service as the medium to display space query and display data, and directly fill in green, yellow, and green on the map The warning colors such as red show the rainfall warning situation, and at the same time, real-time reports can be produced online to provide the central management unit with information about the warning of the whole Taiwan. When the rainfall warning value is exceeded, the system will also directly notify the responsible personnel by mobile phone text message or email, so as to achieve the purpose of instant warning.
Sustainable Impact: Provide real-time flood warning information for 322 townships and districts across Taiwan.
有鑑於近年來都市淹水災害日益嚴重，都市防災示警工作為營建署的重要任務之一。本中心協助建置「都市溢淹示警系統」，以各地區雨水下水道設計標準為基礎，設定下水道系統的雨量警戒值，提供全臺灣322個鄉(鎮、市、區)的都市淹水即時警戒資訊。
本系統主要功能包含現況示警、未來預警、報表、設定、參考資料、CEOC專區、進階管理等；以地圖應用服務為媒介展現空間查詢與展示資料，並直接於地圖上填入綠、黃、紅等警戒顏色呈現雨量警戒情況，同時可於線上產製即時報表，提供中央管理單位瞭解全臺警戒資訊。當超過雨量警戒值時，系統亦以手機簡訊或電子郵件方式直接通知權責人員，以達到即時示警的目的。
永續影響力: 提供全臺322個鄉鎮市區的即時淹水警戒資訊。
|
|
Evidence:
http://www.gis.ntou.edu.tw/AchievementCityflood.html
https://cityflood.nlma.gov.tw/?
- SWMM management platform SWMM檢核管理平臺
In order to implement land information survey and disaster prevention and relief utilization plan, Construction and Construction Department promotes ‘Stormwater Sewer Census’. At the same time, in cooperation with the promotion of stormwater sewer planning and governance, based on the results of the census, the street/sewer SWMM model (including stormwater sewers, street drainage, regional drainage and other urban drainage systems) construction methods are introduced to conduct hydraulic analysis and analysis of the current urban drainage. Review the planning, and then provide value-added applications for urban flood warning.
GIS center assists in proposing the standard operating procedure for the construction of the street/sewer SWMM model, so that the stormwater sewer (review) planning project execution unit can cooperate with it; at the same time, through a series of inspection procedures, confirm the correctness of the water management model, so as to effectively integrate The results of the stormwater sewerage (review) planning projects in various counties and cities. In addition, GIS center has built the ‘SWMM Composite Urban Drainage System Inspection and Management Platform’, which uses cloud technology to provide services such as data transmission, checklist inspection, and inspection progress query through web pages; The correctness of the hydraulic model.
Sustainable Impact: Enhance the application efficiency of urban flood warning systems.
為落實國土資訊調查及防救災運用規劃，營建署推動「雨水下水道普查」。同時配合雨水下水道規劃及治理之推動，以普查成果為基礎，導入街道/下水道SWMM模式(包含雨水下水道、街道排水、區域排水等都市排水系統)建置方式，針對現況都市排水進行水理分析及檢討規劃，進而提供都市溢淹示警的加值應用。
本中心協助提出街道/下水道SWMM模式建置標準作業流程，使雨水下水道(檢討)規劃案執行單位可配合辦理；同時透過一系列的檢核程序，確認水理模式的正確性，以此有效整合各縣市雨水下水道(檢討)規劃案成果。此外，本中心建置「SWMM複合型都市排水系統檢核管理平臺」，利用雲端技術，以網頁方式提供資料傳遞、清單檢查、檢核進度查詢等服務；並提供自主檢查功能，即可自行檢視水理模型的正確性。
永續影響力: 提升都市溢淹示警的應用效能。
Evidence:
http://www.gis.ntou.edu.tw/AchievementSWMM.html
https://inspect.nlma.gov.tw/swmm/?
- SDG14.3Supporting aquatic ecosystems through action SDG14.3 通過行動支援水生生態系統
On a cold winter day in 2015, the European Union issued a troubling warning. At that time, Taiwan's fishing industry faced severe scrutiny—overfishing issues had attracted international concern. A few years later, Taiwan received a ‘‘yellow card’’ for the exploitation of foreign fishermen. These consecutive warnings struck like heavy blows, highlighting the weak points in Taiwan's fisheries management.
In the face of the challenges in fisheries management, finding a solution became urgent. The monitoring system needed an upgrade, particularly in effectively tracking the distribution and status of the large number of vessels. This became a primary objective. However, traditional monitoring systems proved inadequate when dealing with such complex data.
At this critical moment, a group of researchers stepped forward with a vision: to forge a new path for Taiwan’s fisheries. Starting in 2016, we dedicated significant effort and expertise to an important project organized by the Fisheries Agency. We developed an innovative system called the ‘Next-Generation Global 3D Real-Time Dynamic Marine and Fisheries Geographic Information Analysis System.’ This system functioned like a powerful brain, capable of processing vast amounts of data and performing intelligent analysis using AI technology. Its visualization and dynamic features enabled monitoring personnel to easily grasp the status of fisheries and make swift judgments.
But the achievements of this system extend beyond fisheries management. Its applications have progressively expanded to real-time management of distant-water fisheries and even ventured into the field of meteorology. By integrating meteorological data, it not only assesses fisheries and marine resources but also predicts ocean conditions, providing fishermen with more reliable operational guidance. Remarkably, the system's advantages have also extended to the assessment of offshore wind farms and the setting of shipping routes.
The research team's ambitions do not stop here. We hope to integrate satellite technology in the future to further enhance the capabilities of the information analysis system. This innovation will not only benefit domestic operations but also contribute to international traffic and crowd management. Taiwan's intelligence and efforts will shine brightly on the global stage.
Sustainable Impact: Developed the ‘Next Generation Global 3D Real-Time Dynamic Ocean and Fisheries Geographic Information Analysis System,’ enabling monitors to easily grasp fishery conditions and make quick decisions.
在2015年的一個寒冷冬日，歐盟發出了一個令人擔憂的警告。那時，臺灣的漁業面臨著嚴峻的考驗——濫捕問題引起了國際社會的關切。幾年後，臺灣又因為外國漁工的被壓榨問題而收到了「黃牌」。這些連續的警訊像一記記重擊，直指臺灣漁業管理的薄弱環節。
在漁業管理的難題面前，解決方案顯得迫在眉睫。監測系統亟需升級，尤其是如何在龐大的船隻數量中有效掌握其分布與狀況，成為了首要目標。然而，傳統的監測系統在面對如此繁雜的數據時顯得捉襟見肘。
就在這個關鍵時刻，一群科研人員站了出來。我們的心中懷著一個願景，那就是為臺灣漁業開創一條全新的道路。從2016年開始，我們投入了大量的精力和智慧，參與了由漁業署安排的一項重要計畫。我們誕生了一個名為「新世代全球3D即時動態海洋與漁業地理資訊分析系統」的創新系統。這個系統像是一台強大的大腦，能夠處理海量的數據，並通過AI技術進行智能分析。它的視覺化和動態化特性，使得監測人員能夠輕鬆掌握漁業狀況，並迅速做出判斷。
但這個系統的成就不僅僅停留在漁業管理上。它的應用逐步拓展到遠洋漁業的即時管理，並且跨足到氣象領域。結合氣象數據，它不僅能夠進行漁業海況資源的評估，還能預測海洋狀況，為漁民提供了更為可靠的作業指南。更令人驚喜的是，這個系統的優勢還延伸到了離岸風力發電廠的評估和船舶航道設定上。
科研團隊的夢想並未止步於此。我們希望未來能夠結合衛星技術，進一步提升信息分析系統的能力。這樣的創新不僅能夠在國內發揮作用，也將為國際間的交通與人流管理提供貢獻。臺灣的智慧與努力，將在全球舞台上閃耀出一抹亮麗的光彩。
永續影響力: 開發了「新世代全球3D即時動態海洋與漁業地理資訊分析系統」。使監測人員能輕鬆掌握漁業狀況，迅速做出判斷。
|
|
|
Evidence:
https://www.youtube.com/watch?v=UX5Zl5wG4bw
4. The ecosystem dynamics associated with the environmental change surrounding the upwelling zone of Taiwan Bank resources in the coastal water off Taiwan maculate in the southwestern waters of Taiwan台灣淺灘(Taiwan Bank)湧昇區暨周邊水域生態系動態特性影響之研究
The knowledge regarding the feeding ecology of the greater amberjack remains limited despite its ecological and economic importance. Therefore, this study investigated the feeding dynamics of the greater amberjack and its key prey species in Taiwanese waters. Samples were collected from the Taiwan Bank (22.5° N–24.5° N, 118.5° E–121.5° E) and Northern Taiwan waters (25° N–26° N, 121° E–123.5° E). Analysis of samples obtained between June 2020 and June 2022 indicates that environmental factors influenced the availability of prey species, thereby affecting dietary preferences. The prey-specific index of relative importance indicated that the predominant prey species were pelagic species (52.81%), followed by demersal species and unidentified teleosts (18.74% and 16.58%, respectively); squid and crustaceans were the least frequently consumed prey species (6.11% and 5.76%, respectively). Although no difference was discovered between males and females in terms of diet, seasonal and size-related variations were noted in the feeding patterns, as evident from the substantial ontogenetic shift observed in the dietary composition of the samples during the study period. The findings of this study improve the understanding of the feeding dynamics of the greater amberjack and the association between oceanographic factors, such as sea surface temperature (SST), sea surface salinity (SSS), sea surface height (SSH), mixed-layer depth (MLD), and eddy kinetic energy (EKE), and prey availability in the coastal waters of Taiwan.
Sustainable Impact: The results of this study contribute to further exploring the ecological role of the greater amberjack in Taiwan's coastal waters and its relationship with environmental changes.
儘管杜氏鰤具有生態和經濟重要性，但有關其攝食生態的知識仍然有限。因此，本研究調查了台灣水域中杜氏鰤及其主要餌料生物的攝食動態。樣本採集自台灣淺灘(22.5°N-24.5°N, 118.5°E–121.5°E)及台灣北部水域(25°N-26°N, 121°E-123.5°E)。由2020年6月至2022年6月的樣本分析顯示，環境因素影響了餌料生物物種的可用性，從而影響了攝食偏好。餌料生物特異性相對重要性指數顯示，主要餌料生物種類為中上層種類(52.81%)，其次是底層種類和不明硬骨魚(分別為18.74%和16.58%)；魷魚和甲殼類動物是最不常被食用的餌料生物物種(分別為 6.11% 和 5.76%)。儘管在攝食方面沒有發現雄性和雌性之間的差異，但在攝食模式中註意到季節性和與體型相關的變化，這一點從研究期間觀察到的樣本飲食成分中觀察到的實質性個體發育變化中可以明顯看出。這項研究的結果提高了對杜氏鰤攝食動態以及海洋因素之間關聯的理解，例如海面溫度(SST)、海面鹽度(SSS)、海面高度(SSH)）、混合層深度(MLD)、渦動能(EKE)以及台灣沿海水域的餌料生物可用性。
永續影響力：本研究結果有助於進一步探索杜氏鰤在台灣沿海水域的生態角色及其與環境變化的關聯。
|
|
|
Evidence:
https://r088.ntou.edu.tw/var/file/103/1103/img/1564/757150331.pdf
5. 112th ‘‘Advancing Satellite Monitoring Marine Environment Application Products’’ outsourcing case 112年度「精進衛星監測海洋環境應用產品」委外案
This project will make use of the continuous observation characteristics of Sunflower satellite data in high spatial and temporal resolution and panoramic range to refine sea surface chlorophyll concentration and sea surface temperature calculation algorithms, strengthen the display function of the satellite product fishery information platform, and explore the surrounding waters and coastal areas of Taiwan. , for the collection and analysis of disaster-causing weather and fishery (including aquaculture) information, it is expected to establish an early risk estimation model for satellite products in fishery disasters in the context of extreme weather affecting fish habitats, and reduce economic and social losses. risk. The main work of this year (112) is to (1) improve the sea surface chlorophyll concentration algorithm in the panoramic range; (2) develop artificial intelligence methods to establish all-weather hourly sea surface temperature products in target sea areas; (3) apply artificial intelligence to develop PM2 in East Asia .5 Concentration ground estimation field production technology; (4) Improve satellite fishery platform information and operational function optimization; (5) Explore the status of disaster-causing weather and satellite products and fishery (including breeding) information.
In the development of panoramic sea surface chlorophyll concentration algorithm project, ocean water color algorithms were established for three sea areas. The results show that the chlorophyll concentration value of the Sunflower satellite panorama in this study is consistent with MODIS/Aqua, both in value and spatial distribution. The results of the chlorophyll concentration values of MODIS /Terra, SNPP-VIIRS, and JPSS1-NOAA20 are quite consistent. This means that the accuracy of the chlorophyll concentration values and distribution maps calculated by this study on the Sunflower satellite panorama is at a certain level.
In the research and development of artificial intelligence methods to establish all-weather hourly sea surface temperature products in target sea areas, this year’s work expanded the time-spatial-spatial radial basis function neural network (TS-RBFNN) to hourly sea surface temperature products. Surface temperature reconstruction and continuous development of the Temporal Periodic Neural Network (TPNN) model module. Its principle is similar to the developed amplitude basis function neural network. The main difference is that it is modified into a periodic function ( Determine the number of neurons and the period of sea surface temperature changes) to reflect the periodic changes in sea surface temperature. In the actual case of hourly sea temperature reconstruction, the root mean square error (0.436°C) of the reconstructed sea temperature of TPNN is higher than that of Temporal Radial Based Function Neural Network (TRBFNN), and the calculation efficiency is It is about three times faster than TRBFNN (0.0064 seconds).
In the technical project of applying artificial intelligence to develop a ground-based estimation field for PM2.5 concentration in East Asia, a prototype model of a ground-based estimation field for artificial intelligence PM2.5 concentration in East Asia has been initially built. At this stage, the model can effectively grasp the real observed concentration. In the final stage, the 110-year DNN model design was improved. By combining the PM2.5 concentration observation difference mechanism at the measuring station to evaluate the impact on the concentration estimates in coastal and mountainous areas, the estimation results of sea grid concentration in Taiwan and East Asia were provided. .
In the project of upgrading the information and operational functions of the satellite fishery platform, we have completed the refinement of habitat suitability and species distribution model forecasting functions, and planned and constructed the important fishery economic species model for mullet, sea bream, black sea bream, pomfret (white seabream). There are a total of 10 species of economically important fish species along the coast, including pomfret, pomfret, pomfret, and kingfish. The platform information integrates meteorological satellite and hydrological environment data, mainly providing information related to fishermen's operations and fishing ground forecast needs and fishery resource monitoring. Satellite telemetry data can be used as reference and evidence for academic research or project implementation, and through satellite telemetry image analysis, A marine fishery service application platform can be built to provide more diversified and customized service products, improving the application level of satellite products and the convenience and promotion of the operational service system.
In the project to explore disaster-causing weather and satellite products and fishery (including aquaculture) information, this year's analysis mainly focused on Taiwan's important farmed fish species - milkfish. The main production areas are Tainan City, Kaohsiung City, Chiayi County and Yunnan Province. In Lin County, according to the production type, it can be mainly divided into deep water exclusive culture and shallow flat polyculture. Milkfish is a temperate narrow-temperature bony fish with poor cold tolerance. Generally, its activity will significantly decline when the water temperature is lower than 14°C, and individuals will begin to die below 13°C. This study collected more than 110,000 water temperature data from the Tainan Water Quality Monitoring Station (sensor water depth is 40 cm) and temperature data from four nearby weather stations. We used regression analysis and combined with the temperature hazard quotient to analyze the correlation between water temperature and air temperature. The results show that water temperature has a delay characteristic of at least 3 hours with air temperature. When the temperature is 10°C, the water temperature of the breeding pond is about 13°C. Under these conditions, milkfish begin to die, causing breeding losses.
Sustainable Impact: The outcomes of this project not only enhance the efficiency of fishery management but also provide scientific evidence for response strategies under extreme weather conditions, thereby promoting the sustainable development of Taiwan's fisheries.
本計畫將利用向日葵衛星資料在高時空解析度及全景範圍的連續觀測特性，精進海表面葉綠素濃度及海表溫演算法，強化衛星產品漁業資訊平臺展示功能，並探討臺灣周遭海域與沿岸區域，對於致災之天候與漁業(含養殖)資訊的樣態蒐集分析，期能為極端天候影響魚種棲地的情境，建立衛星產品在漁業災害的早期風險推估模式，降低經濟與社會損失的風險。本(112)年度主要工作為(一) 提升全景範圍的海表面葉綠素濃度演算法；(二) 研發人工智慧方法建立目標海域全天候逐時海表溫產品；(三) 應用人工智慧開發東亞區域PM2.5濃度地面推估場產製技術；(四)提升衛星漁業平台資訊及操作功能優化；(五) 探討致災天候及衛星產品與漁業(含養殖)資訊的樣態。
在發展全景範圍的海表面葉綠素濃度演算法工項，針對三種海域分別建立海洋水色演算法，結果顯示本研究向日葵衛星全景的葉綠素濃度值不論在值或空間的分布，都與MODIS/Aqua, MODIS/Terra, SNPP-VIIRS, JPSS1-NOAA20的葉綠素濃度值的結果相當一致。這表示本研究所推算向日葵衛星全景的葉綠素濃度值及其分布圖的準確度有一定的水準。
在研發人工智慧方法建立目標海域全天候逐時海表溫產品工項，本年度工作拓展時間–空間幅狀基底函數類神經網路(Temporal-Spatial Radial Basis Function Neural Network, TS-RBFNN)至逐時海表溫重建及持續開發時間週期性類神經網路模式模組(Temporal Periodic Neural Network, TPNN)，其原理與已開發之幅狀基底函數類神經網路相似，主要差異在改納入週期性函數(決定神經元個數及其海表溫變化之週期)以反應海表溫的週期性變化。在實際變化的逐時海溫重建案例中，TPNN的重建海溫之均方根誤差(0.436℃)高於時間幅狀基底類神經網路(Temporal Radial Based Function Neural Network, TRBFNN)，計算之效率較TRBFNN提升三倍左右(0.0064秒)。
在應用人工智慧開發東亞區域PM2.5濃度地面推估場產製技術工項，初步建置東亞區域人工智慧PM2.5濃度地面推估場模型雛形，現階段模型已可有效掌握真實觀測濃度。於期末階段改良110年之DNN模型設計，藉由結合測站PM2.5濃度觀測差值機制以評估對於沿海與山區濃度推估值的影響，提供臺灣區域與東亞區域海上格點濃度推估結果。
在提升衛星漁業平台資訊及操作功能優化工項，目前已完成精進棲地適合度與物種分布模式預報功能，在重要漁業經濟物種模式規劃建置烏魚、午仔、黑鯛、鯧科(白鯧、定盤)、鰺科與石首魚等，沿近海重要經濟魚種共計10種。平台資訊整合氣象衛星水文環境資料，提供漁民作業及漁場預測需求與漁業資源監控相關資訊為主；衛星遙測資料可作為學術研究或執行計畫時之參考與佐證，並透過衛星遙測影像分析，可建置海洋漁業服務應用平台，提供更多樣化與客製化服務之商品，提升衛星產品應用層面及作業服務系統便利性及推廣使用度。
在探討致災天候及衛星產品與漁業(含養殖)資訊的樣態工項，本年度主要針對臺灣重要養殖魚種-虱目魚進行分析，主要生產區為臺南市、高雄市、嘉義縣及雲林縣，依照生產型態主要可分為深水專養及淺坪混養。虱目魚為溫帶狹溫性硬骨魚類，耐寒性不佳，一般水溫低於14 ℃其活動力便明顯下滑，13 ℃以下便會開始有個體死亡。本研究蒐集台南水質監測站（感應器水深40公分）共逾11萬筆水溫資料與鄰近4個氣象站之氣溫資料利用迴歸分析並搭配氣溫危害商數，進行水溫與氣溫之相關性分析。結果顯示，水溫之於氣溫至少有3小時的延遲特性，而在氣溫10℃時，養殖池塘水溫約為13℃，在此條件下虱目魚便開始產生死亡，造成養殖損失。
永續影響力：本計畫的成果不僅能提升漁業管理的效率，還能為極端天候下的漁業應對策略提供科學依據，進而促進臺灣漁業的可持續發展。
|
|
|
Evidence:
https://r088.ntou.edu.tw/var/file/103/1103/img/1564/804006726.pdf