宁夏生态足迹影响因子的偏最小二乘回归分析

生态足迹分析方法是一种度量区域生态可持续程度的有效方法,偏最小二乘回归法(PLS)能有效解决多元回归分析中变量的多重相关性问题,具有容易操作,相关分析精度高等特点。以宁夏为研究区域,在计算了宁夏2001—2010年人均生态足迹的基础上,应用偏最小二乘回归分析法,对影响宁夏生态足迹的各因子的重要程度进行了分析。通过变量投影重要性分析、特异点分析和预测分析,证明所得偏最小二乘回归模型具有较好的精度。研究结果为:2001—2010年,宁夏人均生态足迹由1.818103793 hm2上升至2.894958909 hm2,生态赤字由1.28352051 hm2上升至2.42316627 hm2,生态承载力由0.53458328 hm2下降至0.47179264 hm2;全区GDP、城镇居民人均生活消费支出、第二产业产值和第一产业产值是影响宁夏生态足迹的显著因子。

Analysis the relationship between ecological footprint (EF) of ningxia and influencing factors: Partial Least-Squares Regression (PLS)

The model of ecological footprint (EF) was originally brought forward by ecological economist William E. Rees and his fellows in 1992, and Wackernage modified and perfected the model in 1996. Ecological footprint is an effective method to evaluate the ecological security and sustainability of a region from the aspects of demand and supply. Ecological footprint is defined as biologically productive areas required to sustainably support certain population in a region, and ecological footprint theory can be simply generalized as follows. First, people's daily consumption should be classified and the biologically productive area needed to support the consumption should be caculated. The ecological footprint of people is the summation of biologically productive area. We can then reckon the ecological capacity and convert to the biologically productive land area for comparison. At last, we can compare the ecologcial footprint with the ecological capacity Partial. Least-Squares Regression (PLS) method was proposed by S.Wold and C.Albano in 1983. As a statistical tool, PLS regression has been specifically designed to deal with multiple regression problems where the number of observations are limited and correlations between variables are high. PLS regression has gained great success in scientific fields, such as chemometrics, medicine, market analysis and finance. There are multi-correlation problems between ecological footprint's influencing factors of multiple regression analysis. PLS regression aims at producing a model that can transform a set of correlated explanatory variables into a new set of uncorrelated variables, called PLS factors in this paper. PLS factors capture most information of the independent variables that is useful for explaining and predicting the dependent variables. In the meantime, PLS regression reduces the dimensionality of the regression by using fewer PLS factors than the number of independent variables. The per capita EF and ecological carrying capacity (EC) of Ningxia are calculated from 2001 to 2010 by employing the quantitative method for EF and major important factors influencing the ecological footprint selected by PLS regression. It is found that ecological footprint of Ningxia increased from 1.818103793 hm² to 2.894958909 hm², ecological deficit (ED) increased from 1.28352051 hm² to 2.42316627 hm² and ecological capacity (EC) decreased from 0.53458328 hm² to 0.47179264 hm². GDP of Ningxia, urban per capita life expenditure, the secondary industry, and the primary industry are the main factors influencing ecological footprint of Ningxia.