ABSTRACT Urban landscapes are changing in response to changes in socio-economic conditions. Land change scientists seek to understand these land dynamics in the coupled human-environment system of urban landscapes. This study assessed land change in the built-up area of Wa Municipality between 1986 and 2016 using Landsat images. We used the Support Vector Machine algorithm for classifying the images. We recorded image classification accuracies of 97%, 95%, 92% and 96% for the 1986, 1996, 2006 and 2016 classified images, respectively. Our study finds that over the 1986–2016 period, agricultural land and bare land transitioned to build-up land by 9.23% and 3.79%, respectively, as compared to 2.79% for vegetation and 0.05% for water. Our in-municipal level analysis thus shows that urban landscapes could expand more sustainably by targeting other dominant land categories instead of the vegetation cover. The findings in this paper could serve as a spatial model for planning and reducing the unintended socio-ecological impacts of expansion in the built-up area.

中文翻译：

---

一种评估加纳瓦市建成区土地变化的地理空间方法

摘要 城市景观随着社会经济条件的变化而变化。土地变化科学家试图在城市景观的耦合人类环境系统中了解这些土地动态。本研究使用 Landsat 图像评估了 1986 年至 2016 年间佤市建成区的土地变化。我们使用支持向量机算法对图像进行分类。我们分别记录了 1986、1996、2006 和 2016 年分类图像的 97%、95%、92% 和 96% 的图像分类准确率。我们的研究发现，在 1986-2016 年期间，农业用地和裸地分别转变为建设用地的 9.23% 和 3.79%，而植被为 2.79%，水为 0.05%。因此，我们的市内水平分析表明，通过瞄准其他主要土地类别而不是植被覆盖，城市景观可以更可持续地扩展。本文的研究结果可以作为一个空间模型，用于规划和减少建成区扩张的意外社会生态影响。