臺灣地區森林覆蓋與用電量關聯性之探討

Abstract

本研究之主旨為探討相關變數對於人均用電量之影響，由於臺灣位在不同地理區域的交界帶上，故地形、地質、氣候與生態等方面具有多樣性的特質，提供多樣化產業發展的可能性，且每個縣市均具有其獨特的自然環境條件、發展定位與產業結構等，造成區域用電量的成因影響因子眾多，將衍生出不同的用電行為。 本研究蒐集 2007~2012 年間臺灣各縣市的人均用電量、森林覆蓋率、人口密度、年均溫、平均每人每年可支配所得、每百戶擁有之冷暖氣機數、每百戶擁有之電腦數及每百戶擁有之電視機數等資料，並應用逐步迴歸分析來選取能夠有意義的解釋變數，發現使用每百戶擁有之冷暖氣機數、每百戶擁有之電腦數、年均溫及森林覆蓋率等4個因子可解釋年人均用電量的影響機制，產生的迴歸模型解釋變異度達0.774，其中又以「每百戶擁有之冷暖氣機數」、「每百戶擁有之電腦數」因子較為重要。因此，日後在進行相關模式研究時，可考慮將這些因子納入，以提升模式預測的準確性。

The objective of this study is to analyze the influence of relevant variables to annual electricity consumption per capita. Taiwan is located at the junction of different geographical regions which bring diversity to topography, geology, climate, ecological characteristics as well as a variety of industrial development possibilities. Each county has its unique natural environment, positioning development and industrial structure that led electricity consumption derived from different behavior. This study collected annual electricity consumption per capita, forest coverage rate, population density, mean annual temperature, mean annual disposable income per capita, numbers of air-conditioning machines, computer and TV owned by every hundred household in Taiwan from 2007 to 2012, and apply stepwise regression procedure to select meaningful independent variables. This study also found the impact mechanism of annual electricity consumption per capita can be explained by numbers of air-conditioning machines and computer owned by every hundred household, mean annual temperature and forest coverage rate. The resulting regression model described 77.4% of variability. Therefore, when someone is doing related research patterns in the future, he/she could consider these factors, in order to enhance the accuracy of regression model.