光電業者對於農地上設置綠能光電之空間依存度分析─以雲林縣為例

農地為農業的重要資產，農地農用與農民收入及國家糧食安全息息相關。雲林縣為台灣農業生產大縣，同時也是農地設置太陽能板最多的縣市。本研究採用雲林縣2018年434筆太陽能板設置資料，探討太陽能板設置於農地上是否具群聚效應，以及影響群聚及設置規模大小的內外關鍵因素。本文首先透過空間分析Moran's I和LISA檢測分析農地太陽能板群聚狀況，再應用分量迴歸進一步找出影響裝設太陽能板的重要因子。空間分析結果顯示，雲林縣的太陽能板設置具有聚集現象。顯著熱點聚集的區域為四湖、臺西、口湖、麥寮等鄉鎮，其中四湖鄉為屋頂型及地面型太陽能板最集中之區域；顯著冷點則為水林、東勢、元長等鄉鎮。分量迴歸結果顯示，公司代表人持股比例每增加1%太陽能板設置面積減少6m2到8 m2，表示代表人持股比越高設置大面積太陽能板的意願越低。老年人口比例每增加1%設置面積則下降100 m2到150 m2。顯著影響太陽能板設置面積的因素還包括所營事業、公司關聯性、至市中心之距離、可耕地面積比例等等。建議政府以不影響農地農用為前提發展再生能源，可透過限制農地設置太陽能板的面積，並將地面型17GW的發展目標轉向發展屋頂型設施。

關鍵字

綠能光電；農地；群聚效應；空間自相關分析；分量回歸

並列摘要

Land is the most important facet of agriculture. Yet, the aspect of agricultural land is intimately related to farmers’ income and food security. In Taiwan, Yunlin County has the largest agricultural land. Meanwhile, the County has the highest installations of photovoltaic panels on farmland. This study seek to map and investigate the determinants of photovoltaic panel deployment in Yunlin. Using 434 satellite data of Yunlin County for the year 2018, the study explore the clustering effect and the determinants of Photovoltaic panel variability. The study applied Moran's I and LISA method, to assess the deployment of the green energy facilities. Furthermore, the study applied quantile regression to examine the impact of photovoltaic panel variability in Yunlin. Empirical results revealed that the distribution of the green energy facilities in Yunlin County is clustered, with Sihu, Taixi, Kouhu, and Mailiao as the hotpots. Explicitly, the study showed that the installation of photovoltaic is significantly large in Sihu Township. In contrast, Shuilin, Dongshi and Yuanchang were identified as the cold spot of photovoltaic panel installation. Moreover, quantile regression results indicate that shareholding ratio increase 1% and panels area will decrease 6m2 to 8m2, proportion of aged population increase 1% and panels area will decrease 100 m2 to 150 m2, factors that significantly affect the installation area of solar panels also include business items, commercial farming, distance from city center, the ratio of arable land, are among the factors that affect the induction of photovoltaic panels in Yunlin County. Through the analysis, it is recommended that the government should switch to a rooftop solar photovoltaic panel system. Although the roof-type installation area is smaller and risky than the ground type, it does not affect the use of agricultural land and can increase farmers' income and stimulate the development of renewable energy.