| DC 欄位 |
值 |
語言 |
| DC.contributor | 土木工程學系 | zh_TW |
| DC.creator | 魏霈萱 | zh_TW |
| DC.creator | Pei-Hsuan Wei | en_US |
| dc.date.accessioned | 2020-1-22T07:39:07Z | |
| dc.date.available | 2020-1-22T07:39:07Z | |
| dc.date.issued | 2020 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=106322089 | |
| dc.contributor.department | 土木工程學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 氣候變遷導致的自然災害，影響農民收益來源且引發糧食安全問題，政府為分擔農業生產風險、穩定社會經濟情勢，積極地推動農業保險。然而，現階段的農業保險機制未臻完善，導致農民投保意願偏低，且至現地進行損失及給付評估，耗費保險公司大量時間與人力成本，為農業保險推展之瓶頸。本研究期望結合遙感探測技術與數據分析，改善理賠判釋作業耗費時間與人力成本的困境，輔助農業天然災害保險之設計，提高保險合約的可靠性與效益。
本研究試驗區域位於臺灣中西部的雲林縣，屬於亞熱帶氣候，其稻米栽種面積廣泛、品質優良，為主要的經濟農業作物。此研究運用遙測科技擴大偵測範圍，分析地球資源衛星Landsat影像計算物候參數，常態化差異植生指數（Normalized Difference Vegetation Index, NDVI）、綠波段常態化差異植生指數（Green Normalized Difference Vegetation Index, GNDVI），提供農業生態系統的空間與時間訊息，瞭解目標區域之稻米生育變化，並且蒐集中央氣象局之氣候資訊，如生長度日（Growing Degree-days, GDd）、降雨量、風速等作為環境指標。將2008至2017年之數據導入迴歸分析之統計模型，於稻米生長週期進行產量預測。具體分析作物產量之影響因子，探討衛星資訊納入於農業保險機制之可行方向。
研究結果顯示，透過植生指數的時間序列可以得知水稻的生長狀況與物候期，而利用綠波段常態化差異植生指數建立的線性回歸模型，具有最佳的產量預測能力。此外，最適宜產量估算之物候期為生殖階段(reproductive phase)，並且第一期稻作之預測誤差低於第二期稻作。然而，由於農民之間耕作方法存在差異，且有不利的天氣狀況，皆會導致物候參數混亂，進而影響產量估算的效果。考量水稻之物候特徵型態，植生指數於作物成熟階段(ripening phase)擁有輔助災損評估之潛力。總體而言，遙測技術應用於農作物之生長監測與產量估算具備可行性，且提高農業保險之成本效益，穩定糧食安全風險。 | zh_TW |
| dc.description.abstract | As the increased chance of extreme weather events, agricultural damages induced by natural hazards not only affect farmers’ revenue but also threaten food security. In order to absorb the risk of production as well as stabilize the economy for society, Taiwan’s government urges to develop an agricultural insurance program. However, the imperfections of the existing insurance contracts encounter low willingness of participants in the pool. For assessing the actual losses and determining the insurance benefits, the insurance company has to conduct extensive and intensive in-field measurement of yield in the entire areas, which need a lot of time and labor costs. The purpose of this study is to develop a prototype for yield forecast and loss assessment of rice, integrating remote sensing technique and data analytics into the assessment system to improve efficiency and reduce capital expenditure in agricultural insurance.
The study area situates Yunlin County in the central part of western Taiwan and belongs to the subtropical climate zone. Rice is an important economic crop known for its high quality in this area. A large-scale observation is applied from remote sensing that provides the spatial and temporal information of the agroecosystem. This research analyzes the Normalized Difference Vegetation Index (NDVI) and Green Normalized Difference Vegetation Index (GNDVI) by monitoring the phenological information from the U.S. Geological Survey’s Landsat satellite imagery and collects meteorological parameters from the Central Weather Bureau of Taiwan, including Growing Degree-days (GDd), precipitation, and mean wind speed. The yield estimation models are established by multiple regression analysis with the historical data from 2008 to 2017. In addition, the impact factors of production are explored and conceived for agricultural insurance mechanism.
Based on the results, the growth condition and phenological stage of rice could be depicted from the temporal profiles of vegetation indices. The linear regression model of GNDVI has the outstanding ability for estimating rice yield. Moreover, the appropriate stage for yield prediction is the reproductive phase. The estimated result of the second crop is worse than the first crop. Inconsistent cultivation practice and unfavorable weather information could cause cluttered phenological information and influence the efficacy for yield estimation. Considering the phenological characteristics of rice, vegetation indices have the potential to assist loss assessment during the ripening phase. This study concludes that remote sensing is feasible for crop monitoring and yield estimation as well as cost-effective for agricultural insurance, enduring food security by mitigating external risks. | en_US |
| DC.subject | 地球資源衛星 | zh_TW |
| DC.subject | 產量預測 | zh_TW |
| DC.subject | 水稻 | zh_TW |
| DC.subject | 臺灣 | zh_TW |
| DC.subject | Landsat | en_US |
| DC.subject | Yield Forecast | en_US |
| DC.subject | Paddy Rice | en_US |
| DC.subject | Taiwan | en_US |
| DC.title | 利用衛星遙測資訊於稻米產量預測 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Yield Forecast of Paddy Rice from Satellite Observations | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |