Objective  Cities are a combination of "gray space" and "green space", consisting of both man-made and natural environments. The massive expansion of urban built-up areas and the explosive growth of urban population in the past decades have caused negative impacts on the environment, economy and society. As a component of urban green space, urban parks play a crucial role in the sustainable development of cities. However, the spatial layout of urban parks is not always coupled with population distribution, which can lead to a certain degree of inequity in access to parks for residents living in different areas, and the disadvantaged groups in cities are more likely to be deprived of green space. Ensuring the equal accessibility and layout of urban parks is a major challenge in today's urban development, and a key to achieving environmental justice. However, most of the previous researches in China only measure the fairness of park distribution by accessibility level, but fail to consider the fairness indicators comprehensively, not to mention the lack of effective accessibility measurement methods. 2SFCA is the fairest accessibility measurement model that operates by calculating the supply-demand ratio of public facilities, but the underlying algorithm thereof only focuses on the improvement of a single dimension, and the thinking dimension is not broad enough. In view of this, this research considers several key dimensions of the 2SFCA algorithm that can be improved to further perfect the traditional 2SFCA model to measure the accessibility of urban parks in the central urban area of Nanchang, and then analyze the fairness of the spatial layout of urban parks.

Methods  This research plans to use a variety of mathematical models to evaluate the accessibility of urban parks in Nanchang and the equity of their layout. 2SFCA is most widely used in previous researches in terms of accessibility and equity, but the model only considers a single dimension, making the analysis results thereof less accurate. Therefore, this research improves 2SFCA by comprehensively considering the 4 dimensions of decay function, search range, travel pattern, and competition for demand or supply, and adopts the improved model to calculate the park accessibility for human settlements in Nanchang. After that, the research introduces Lorenz curve, Gini index, bivariate local Moran's I and other equity indicators to further analyze the distribution equality of urban park resources in Nanchang, and adopts data analysis software to visualize and analyze relevant data.

Results  From the above analysis, the research findings are as follows. 1) The level of park accessibility varies significantly between different areas in Nanchang. Specifically, the areas with better park accessibility are distributed in the northeast and southwest of the city, generally in the new urban area, or along large natural open spaces such as large rivers and lakes with better ecological background conditions and more new parks and large parks; the areas with poorer park accessibility are distributed in the southeast and northwest of the city, mostly in the old urban area. 2) The spatial layout of park accessibility is generally in a state of inequity, and there is a mismatch between park resources and population density, with such mismatch being especially serious in the center of the old urban area. On the other hand, the limited supply of land and high land price in the old urban center restrict the construction of large parks, which are the reasons for the lack of services provided by parks in the center of the old urban area, while the suburban areas are ideal for large parks due to their sparse population, good natural environment, low land price and sufficient land supply. However, poor public transportation, high travel time cost, and insufficient park facility in suburban areas reduce the attractiveness and utilization rate of the large parks therein. 3) Increasing park service capacity and reducing transportation travel cost can both improve park equity, but the latter is more limited than the former in equity improvement, and increasing park service capacity by about 15% from its original base can best improve park equity.

Conclusion  As the spatial distribution of urban parks in Nanchang still shows variability and inequity, efforts should be made in the following aspects. 1) Adopt differentiated optimization measures in combination with the spatially divergent characteristics of accessibility levels to, by combining the concepts of urban greenways and stamped green areas, creating urban ecological patches using small green areas and unused spaces, and further creating urban ecological patches along existing rivers, lakes and other areas rich in park resources, improve the fairness of urban park accessibility and distribution and solve the problem of unequal distribution of urban parks. 2) Improve park service capacity to a limited extent. By increasing the number of parks within the city or renovating existing parks, the use of park resources can be made more equitable, but research has found that the enhancement effect is greatly reduced when the area of hard-surface park lands is increased beyond a certain limit. Therefore, the best performance can be achieved by renovating the existing parks and keeping the area of new hard-surface park lands within 15%. 3) Reduce the cost of travel between settlements and parks. The accessibility of parks under all four-time threshold scenarios is at a low level, and the distribution is relatively unequal. Therefore, the construction of urban transport infrastructure should be accelerated, the transport linkage between old and new urban areas should be strengthened, and the operational efficiency of the urban transport system should be improved.