站点可达性方法对公交不平等测度的影响研究

基于缓冲区法(Buffer method)的站点可达性是公交不平等测度的基础数据，通常按乘客到站距离计算方式(直线或沿路网)或数量分布方式(随距离衰减或不衰减)对计算站点可达性的直线缓冲区法进行改进，但可达性计算结果差异对公交不平等测度的影响尚不明确。基于此，以浙江省海宁市为例，将直线缓冲区法与路网缓冲区法、直线衰减法、路网衰减法3种改进方法计算的站点可达性与公交基尼系数进行比较。结果发现：按到站距离计算方式(即路网缓冲区法)改进直线缓冲区法使基尼系数降低 13.60%，按乘客数量分布方式改进(即直线衰减法)使基尼系数降低 35.75%，按两种方式共同改进(即路网衰减法)使基尼系数降低40.56%；改进方法均使小区间可达性值差距减小，按乘客数量分布比按到站距离方式的改进效果更明显；市区比农村的公交基尼系数受可达性方法影响更大，这可能与路网、人口分布的城乡差异有关，且不同可达性方法测度城市与农村公交不平等时存在不同甚至相矛盾的结论。研究表明：简化方法会高估公交不平等且对城市与农村影响不同，因人口或路网数据缺失而选择缓冲区法时应核减公交基尼系数值，数据完备时应选用路网衰减法。

Impacts of Accessibility to Transit Measures on Inequality Index of Public Transport

The buffer method of measuring bus stop accessibility is the basis to calculate inequity index of public transport. There are two commonly used ways to improve the buffer method: one is estimating the walking distance that passengers take to a bus stop using either a straight-line or passenger walking routes; the other way is estimating the population with a decay distribution around bus stop. The impact of the improvement of accessibility measures on the value of inequality index is still unclear. Using Haining City of Zhejiang Province as an example, this study compared Gini coefficient of linear buffer method with Gini coefficients of three improved accessibility measurements including the road network buffer method, linear decay method, and road network decay method. The Gini coefficient of linear buffer method decreases 13.60% when it is calculated with the improved walking distance estimation (Road network buffer method). The Gini coefficient decreases by 35.75% with the improved passenger distribution estimation (Straight-line decay method). The Gini coefficient decreases by 40.56% with the improved walking distance and passenger distribution estimations (Road network decay method). The improvements all lead to a reduction in the gap in accessibility values between Traffic Analysis Zones (TAZ)s, and the distribution of passengers is more effective than this improvement in the walking distance. The Gini coefficient in rural areas is affected more by the improved measurement than that in urban area, which might be related to different road network density and population distribution in urban and rural areas. Different inequality indices even conflicting conclusions of the equity of public transport between urban and rural areas can be obtained with different accessibility measures. The study shows that the simplified accessibility measure overestimates inequality index and has different impacts on urban and rural areas. The public transport Gini coefficient should be considered with the buffer method in the absent of road network and population data, and the network decay method is recommended when there are more available data.