Zhang Wenting, Huang Bo
Collage of Resources and Environment, Huazhong Agricultural University, Wuhan, China.
Environ Sci Pollut Res Int. 2015 Mar;22(6):4475-90. doi: 10.1007/s11356-014-3454-y. Epub 2014 Oct 15.
Soil erosion has become a pressing environmental concern worldwide. In addition to such natural factors as slope, rainfall, vegetation cover, and soil characteristics, land-use changes-a direct reflection of human activities-also exert a huge influence on soil erosion. In recent years, such dramatic changes, in conjunction with the increasing trend toward urbanization worldwide, have led to severe soil erosion. Against this backdrop, geographic information system-assisted research on the effects of land-use changes on soil erosion has become increasingly common, producing a number of meaningful results. In most of these studies, however, even when the spatial and temporal effects of land-use changes are evaluated, knowledge of how the resulting data can be used to formulate sound land-use plans is generally lacking. At the same time, land-use decisions are driven by social, environmental, and economic factors and thus cannot be made solely with the goal of controlling soil erosion. To address these issues, a genetic algorithm (GA)-based multi-objective optimization (MOO) approach has been proposed to find a balance among various land-use objectives, including soil erosion control, to achieve sound land-use plans. GA-based MOO offers decision-makers and land-use planners a set of Pareto-optimal solutions from which to choose. Shenzhen, a fast-developing Chinese city that has long suffered from severe soil erosion, is selected as a case study area to validate the efficacy of the GA-based MOO approach for controlling soil erosion. Based on the MOO results, three multiple land-use objectives are proposed for Shenzhen: (1) to minimize soil erosion, (2) to minimize the incompatibility of neighboring land-use types, and (3) to minimize the cost of changes to the status quo. In addition to these land-use objectives, several constraints are also defined: (1) the provision of sufficient built-up land to accommodate a growing population, (2) restrictions on the development of land with a steep slope, and (3) the protection of agricultural land. Three Pareto-optimal solutions are presented and analyzed for comparison. GA-based MOO is found able to solve the multi-objective land-use problem in Shenzhen by making a tradeoff among competing objectives. The outcome is alternative choices for decision-makers and planners.
土壤侵蚀已成为全球紧迫的环境问题。除了坡度、降雨、植被覆盖和土壤特性等自然因素外，土地利用变化——人类活动的直接反映——也对土壤侵蚀产生巨大影响。近年来，这种巨大变化，再加上全球城市化趋势的加剧，导致了严重的土壤侵蚀。在此背景下，利用地理信息系统辅助研究土地利用变化对土壤侵蚀的影响变得越来越普遍，并产生了一些有意义的结果。然而，在大多数这些研究中，即使评估了土地利用变化的时空效应，通常也缺乏如何利用所得数据来制定合理土地利用规划的知识。同时，土地利用决策受到社会、环境和经济因素的驱动，因此不能仅以控制土壤侵蚀为目标来做出决策。为了解决这些问题，已提出一种基于遗传算法（GA）的多目标优化（MOO）方法，以在包括控制土壤侵蚀在内的各种土地利用目标之间找到平衡，从而实现合理的土地利用规划。基于GA的MOO为决策者和土地利用规划者提供了一组帕累托最优解以供选择。深圳是中国一个快速发展的城市，长期遭受严重的土壤侵蚀，被选为案例研究区域，以验证基于GA的MOO方法在控制土壤侵蚀方面的有效性。根据MOO结果，为深圳提出了三个多重土地利用目标：（1）尽量减少土壤侵蚀；（2）尽量减少相邻土地利用类型的不兼容性；（3）尽量降低改变现状的成本。除了这些土地利用目标外，还定义了几个约束条件：（1）提供足够的建设用地以容纳不断增长的人口；（2）限制陡坡土地的开发；（3）保护农业用地。提出并分析了三个帕累托最优解以供比较。结果发现，基于GA的MOO能够通过在相互竞争的目标之间进行权衡来解决深圳的多目标土地利用问题。其结果为决策者和规划者提供了替代选择。
