對於生態保育及管理而言，棲地模式和生物分布圖是非常重要的工具。但是，野生動物分布資訊不但不完整、稀少，而且不確定性很高。萬一模式中的資訊與實際狀況有落差，不但可能會形成錯誤的評估結果與錯誤的規劃，造成資源的浪費，而且還可能引起反效果，傷害欲保育的對象。本研究為「發展自願性地理資訊的空間本體論及其時空間分析之理論與方法—以臺灣鳥類棲地保育為例」的一個子計畫，總計畫目標擬以鳥類研究為例，提出一架構來發展自願性地理空間資訊之空間知識建構技術。本子計畫預計以兩年的時間，處理自願性地理空間資訊所產生鳥類分布圖中資訊不確定性的問題。第一年將利用中華民國鳥會網站所提供 1970 年代至2010 年之間共計30 多年之鳥類調查資料，針對鳥類資料本身的不確定性加以分析與處理，並運用分區比重法解決鳥類資料繪圖單位不確定性的問題，產生信度與效度均較高的空間分布機率圖。第二年則運用第一年的研究成果，整合子計畫一、二、三和四所產生的資訊，估算出各指標性鳥類於不同時期的空間分布機率圖，並整合子計畫五所產生的網路圖，利用空間規畫手法產生生態保育與管理規劃策略，並與原本解析度較粗且範圍不明確的鳥類分布資料所產生的結果進行比對，以突顯本計畫的價值。
Habitat models and species distribution maps are important tools for biological conservation and management. However, information on the spatial distribution of wild animals is usually incomplete, rare, and contains high uncertainty. When the information in habitat models differ from actual conditions, it may result in erroneous assessment and planning, cause waste of resources, and worse, may even harm the species targeted for conservation. This project is a sub-project of “Developing theories and approaches of geospatial ontology and its spatiotemporal analysis from volunteered geographical Information”. The purpose of the major project is to develop a framework and set of development tools for the application of volunteered geographic information, using volunteered bird geographic information. The purpose of this sub-project is to deal with the uncertainty problem in producing bird distribution maps from volunteered geographic information. During the first year, volunteered survey data from the Chinese Wild Bird Federation from 1970 to 2010 will be used to develop a method for analyzing and treating the uncertainty in the bird data, and dasymetric mapping methods will be used to treat the uncertainty in the location of the bird distribution data. The product will be a model to produce avian spatial distribution maps with improved reliability and validity. During the second year, results from the first year, along with supplementary results from companion sub-projects, will be used to produce spatial distribution probability maps for selected indicator species, which in turn, will be used to produce physical conservation and management plans and strategies. These conservation and management plans will then be compared with those produced from the untreated data, to highlight the value of this project.
