Logo

Agriculture ministry is using AI in landslide warning systems

Agriculture ministry is using AI in landslide warning systems

2026-07-22

Taiwan is characterized by its high mountains and deep valleys, where even a torrential downpour can trigger landslides that can potentially bury an entire village. The Agency of Rural Development and Soil and Water Conservation has been introducing AI to its disaster prevention systems with an aim of offering early warnings for landslides. Let’s find out more.

In 2009, Typhoon Morakot struck Xiaolin Village in Jiaxian District, Kaohsiung. A massive landslide wiped out the village, burying more than 470 people alive.

Taiwan, with its high mountains and deep valleys, has a fragile geology. Coupled with frequent rainfall and human development, landslides often lead to major disasters.

Huang Chuang-hsi
Forecaster
Taiwan has high and steep mountains, which, combined with our weather system, easily leads to these kinds of compounding disasters.

Chen Chen-yu
Ministry of Agriculture
We hope to utilize emerging technologies and information power to integrate disaster prevention data more quickly.

In recent years, the Agency of Rural Development and Soil and Water Conservation has introduced AI into its disaster prevention systems, setting up early warning and monitoring systems for debris flows and mudflows.

Chen Chen-yu
Ministry of Agriculture
The primary method is to use rainfall levels for early warnings. We set different critical rainfall thresholds for different areas based on their local geological conditions and how the terrain has evolved.

Based on meteorological data and assessment, if necessary, an emergency response team can be set up to evaluate the situation based on real time data and make a decision on issuing a landslide warning.

Huang Hsiao-yu
ARDSWC, Ministry of Agriculture
The capture rate (of mudflows) is only a little over 10%. We hope that by integrating AI and on-site monitoring equipment, we can issue warnings earlier and better adjust our forecasting mechanisms

Taiwan faces a long-term challenge of extreme weather. By combining technology with human cooperation, we can stand steadier and safer in every future storm, ensuring such tragedies are never repeated.

For more Taiwan news, tune in:
Mon to Fri at 9:30 pm on Channel 152
Tue to Sat at 1 am on Channel 53

#extremeweather #landslides #Taiwanweather #Taiwantyphoon #Taiwandisasters

提前看到土石流前兆 水保署打造"AI泥流預警系統"

2026-07-22

台灣山高谷深,一場豪雨就足以引發山坡崩塌,瞬間掩埋整個村落。如今強颱巴威下周可能海陸警齊發。近年來,農業部農村水保署,將AI引進防災系統,透過大數據與感測器監測,並配合氣象情資,試圖看見土石流危險的前兆,為國人的生命安全把關。

2009年八八風災,高雄甲仙的小林村遭大規模土石流滅村,活埋超過470人。

山高谷深的台灣,由於地質脆弱,加上降雨頻繁、人為開發,土石流一旦發生,往往釀成重大災情。

[[氣象署氣象預報中心副主任 黃樁喜]]
“(台灣)山又高,然後很陡,配上我們的天氣系統,很容易造成這種複合性的災害”

[[農村水保署減災監測組組長 陳振宇]]
“我們就希望說,是不是能夠透過一些,不管是新興的科技,或資訊的力量,能夠把這樣的一個防災的資訊,能夠做一個更快的整合”

水保署近年將人工智慧引進防災系統,建立土石流與泥流的災害預警與監測。

[[農村水保署減災監測組組長 陳振宇]]
“主要是透過雨量的方式,來做一些預警,我們會針對不同地方的這種地質條件,演化的情形,我們會設定不同的警戒雨量”

配合氣象情資,經研判,如有必要,就會成立緊急應變小組,根據即時資訊進行評估,做出發布土石流警戒的決定。

[[農村水保署減災監測組副組長 黃效禹]]
“(泥流)捕捉率大概只有一成多一點左右,我們希望說現在又有AI的話,又有包括現地儀器的監測,能夠讓我們更提早,或者是更來調整我們的警戒預報的一個機制”

台灣正面臨極端氣候的長期挑戰,透過科技與人攜手,在未來可能面臨的每一場風雨中,站得更穩、更安全,不讓悲劇重蹈覆轍。

更多新聞內容,請鎖定:
民視台灣台(152頻道)週一至週五晚上9:30
民視新聞台(53頻道)週二至週六凌晨1:00

Related News

50,000 seat New Taipei Dome will be city’s 4th urban center

2026-07-22

Woman finds new life in professional sports, career after paralyzing condition

2026-07-22

Street food vendor brings Japanese robatayaki to Kaohsiung

2026-07-22