Abstract Google Earth Engine (GEE) has been increasingly used in environmental and urban studies due to its cloud-based geospatial processing capability and accessibility to a large collection of geospatial datasets like Landsat, Modis, etc. However, at present, ecological and urban modeling efforts based on GEE are facing three grave challenges: current illustrations of GEE are to a large extent “straightforward mapping” applications; technical complexities that ecological or urban modelers have to overcome in order to effectively and easily use GEE to develop image processing based environmental models; and the majority of ecological and urban modelers are not aware of new analytical approaches that are becoming available because of the unprecedent geospatial processing capability and large collection of big geospatial datasets GEE has brought to them. The great potential of GEE to support ecological and urban modeling is less explored. In this study, we augmented GEE functions with a few sets of user-customized functions for improving image classification accuracy, estimating ecosystem services, and modeling urban growth sustainability. The paper is the first effort of modeling urban sustainability based on the concept of ecosystem service value (ESV) and in the cloud with GEE; is the first application of classifying GEE Landsat time-series images to compute yearly ESV; and creates the first set of cloud tools to augment GEE for ecologists and urban modelers to model urban sustainability from GEE and ESV. The paper also chose Hohhot City, Inner Mongolia as a case study to model urban sustainability in a time-series 12 years (2005–2016). The case study successfully estimated ecosystem service values and analyzed urban growth sustainability. It also revealed spatial disparities and temporal dynamics of urban growth sustainability in Hohhot City. The study provides an easy-to-adapt illustration on using GEE for image-based ecological and urban modeling.

中文翻译：

---

通过增强 Google Earth Engine (GEE) 在云中模拟城市增长的可持续性

摘要 谷歌地球引擎（GEE）由于其基于云的地理空间处理能力和对大量地理空间数据集（如 Landsat、Modis 等）的可访问性，越来越多地用于环境和城市研究。然而，目前，生态和城市建模基于 GEE 的努力面临三个严峻挑战：目前 GEE 的插图在很大程度上是“直截了当的映射”应用；为了有效且轻松地使用 GEE 开发基于图像处理的环境模型，生态或城市建模者必须克服技术复杂性；大多数生态和城市建模者都没有意识到新的分析方法正在变得可用，因为 GEE 为他们带来了前所未有的地理空间处理能力和大量大型地理空间数据集。GEE 在支持生态和城市建模方面的巨大潜力鲜为人知。在这项研究中，我们用几组用户自定义函数来增强 GEE 函数，以提高图像分类精度、估计生态系统服务和模拟城市增长的可持续性。该论文是首次基于生态系统服务价值 (ESV) 概念和 GEE 在云中对城市可持续性进行建模；是第一个对 GEE Landsat 时间序列图像进行分类以计算年度 ESV 的应用程序；并为生态学家和城市建模者创建了第一套云工具来增强 GEE，以便从 GEE 和 ESV 对城市可持续性进行建模。本文还选择内蒙古呼和浩特市作为案例研究，对 12 年（2005-2016 年）时间序列中的城市可持续性进行建模。该案例研究成功估算了生态系统服务价值并分析了城市增长的可持续性。它还揭示了呼和浩特市城市增长可持续性的空间差异和时间动态。该研究为使用 GEE 进行基于图像的生态和城市建模提供了一个易于调整的说明。该案例研究成功估算了生态系统服务价值并分析了城市增长的可持续性。它还揭示了呼和浩特市城市增长可持续性的空间差异和时间动态。该研究为使用 GEE 进行基于图像的生态和城市建模提供了一个易于调整的说明。该案例研究成功估算了生态系统服务价值并分析了城市增长的可持续性。它还揭示了呼和浩特市城市增长可持续性的空间差异和时间动态。该研究为使用 GEE 进行基于图像的生态和城市建模提供了一个易于调整的说明。