從臺灣傳統溫室建構智慧溫室物聯網系統:以溫室番茄的氣候風險為例

隨著農業 4.0 時代的到來，傳統溫室農業結合物聯網、大數據等技術提升競爭力勢在必行，然而，現行市面上智慧溫室系統相關的產品或案例都是需投入大量資金建設的裝置，較難被普遍在臺灣使用傳統溫室的農民所接受;另外，面對氣候變遷下層出不窮的極端氣侯事件，傳統溫室農業同時面臨著不斷增長的氣候風險，但現行氣候風險評估工具大多知識門檻較高且難以客製化評估個別溫室的氣候風險，導致其相關工具、資訊對於農民而言參考價值有限，因此本研究建立一套以臺灣傳統溫室為基礎的物聯網系統，將氣候風險評估模式(Greenhouse Crop Climate Risk early Warning system, GCCRW)整合其中提供氣候風險預警服務，同時作為一個物聯網平台此系統具有良好的擴充潛能能夠支持將來各種服務需求及相關設備建置。本研究制定一套物聯網系統建置流程作為基準，從案例農場溫室出發深入了解現況條件，而後使用氣候風險模板分析關鍵議題「溫室番茄耕作面臨的氣候風險」，訂定「風險評估物理模式」，後續研究根據風險評估物理模式之內容，分為物聯網系統硬體設備建置與風險評估模式 GCCRW 建置兩部分，兩者之設計相互搭配並於最後整合運作。物聯網系統硬體設備建置中考量實際運作下環境監測、資料計算、儲存、網路通訊等需求完成各項功能;GCCRW 模式建置中整合即時、一週未來、季長期、氣候變遷四個時間尺度的氣象、氣候資料，使用 WGEN、AgriHydro 中的 Multi-WG、GWLF、SDmodel 等模式評估各時間尺度下的氣候風險資訊。其中，季長期時間尺度評估未來三個月的氣候發展狀況及風險;氣候變遷時間尺度則是評估從現在到 2060 年之氣候及風險變化趨勢，以 20 年為一期，每期間隔為 10 年進行分析。驗證結果顯示各時間尺度模式產製之風險資訊合理且適合目標溫室所採納。研究最後說明 GCCRW 模式整合於案例物聯網系統的運作成果，並提出未來發展中可持續優化的項目作為建議。

關鍵字

智慧農業；物聯網；氣候風險評估；氣候變遷；預警

並列摘要

With the advent of the agriculture 4.0 era, proposed by Council of Agriculture, Executive Yuan, it is imperative for traditional greenhouse agriculture to combine technologies such as the Internet of Things and big data to enhance competitiveness. However, most of smart greenhouse system-related products on the market or cases under studying currently require lots of investment in construction, which is difficult to be accepted by farmers who generally use traditional greenhouses in Taiwan. In addition, in the face of endless extreme weather events under climate change, traditional greenhouse agriculture also faces increasing climate risks, but most of the relevant tools and information nowadays are worthless for farmers because of the high knowledge thresholds and the difficulty of customization. Therefore, this study proposed an Internet of Things system based on Taiwan’s traditional greenhouses and integrated the climate risk assessment model (Greenhouse Crop Climate Risk early Warning system, GCCRW) to provide climate risk early warning services. The system was designed to be easily expandable, and could support various service requirements with related equipment construction in the future. An “IoT system construction process” is developed as a benchmark, starting from the case study of target greenhouse, and then using the Climate Risk Template to analyze the key topic "climate risks faced by greenhouse tomato farming", and following formulating a "risk assessment physical model". According to the risk assessment physical model, the follow-up research is divided into two parts: the hardware equipment construction of the IoT system and the construction of the risk assessment model GCCRW. Operational demands such as environmental monitoring, data calculation, storing, and network communication are considered in the construction of hardware equipment of the IoT system while multiple time scales are designed in GCCRW model in order to differentiate the methodology in proper way, real-time, one week in the future, seasonal long-term and climate change included. Among them, the seasonal long-term time scale evaluates the climate risks in the next three months, and the climate change time scale evaluates the trends of climate risk from now to 2060, with a period of 20 years and an interval of 10 years. In the methodology, various sub-models including WGEN, Multi- WG, GWLF, and SDmodel of AgriHydro are used to evaluate climate risk information. Overall, risk information produced by the system is reasonable and suitable for the target greenhouse of every time scale. At the end of the study, the operation results of the GCCRW model integrated in the case of the IoT system are presented with future development goals as suggestion.

並列關鍵字

Smart Agriculture ； Internet of Things ； Climate Risk Assessment ； Climate Change ； Early Warning

參考文獻

1. Ahmed, K. F., Wang, G., Silander, J., Wilson, A. M., Allen, J. M., Horton, R., Anyah, R. (2013). Statistical downscaling and bias correction of climate model outputs for climate change impact assessment in the US northeast. Global and Planetary Change, 100, 320-332.