ABSTRACT

Melbourne is one of the most livable cities in the world and it is expected to become a global city of opportunities by 2050. These features along with some of the challenges Melbourne will face with in the future, necessitate land use change simulations. Therefore, land use changes in north Melbourne with an area of 2,425.42 km² during 2014–2019 and its future changes until 2030 and 2050 have been simulated. Landsat 8 Operational Land Imager (OLI) and Multilayer Perceptron (MLP) neural networks were used along with Markov chain model in ArcGIS and TerrSet software. Results showed that while proportions of residential and industrial–commercial land uses increased from 35.90% (870.70 km²) in 2014 to 38.53% (934.50 km²) in 2019, that of forest and agricultural-grassland decreased from 62.86% (1,524.01 km²) in 2014 to 57.76% (1,400.99 km²) in 2019. Similarly, the simulation results show that residential and industrial–commercial land use will increase to 42.86% (1,037.139 km²) and 44.53% (1,079.99 km²) by 2030 and 2050, respectively. In the same period, forest and grassland-agricultural land uses are respectively expected to decline from 53.53% (1298.53 km²) to 51.76% (1255.49 km²). Spatial changes will occur mostly in the north and northwest of the Melbourne.

摘要

墨尔本是世界上最宜居的城市之一，预计到2050年将成为一个充满机遇的全球性城市。这些特征以及墨尔本未来将面临的一些挑战，必须进行土地利用变化模拟。因此，模拟了北墨尔本北部2014-2019年期间的土地利用变化，其面积为2,425.42 km ²，并且模拟了直至2030年和2050年的未来变化。在ArcGIS和TerrSet软件中，将Landsat 8操作性陆地成像仪（OLI）和多层感知器（MLP）神经网络与Markov链模型一起使用。结果显示，虽然住宅和工商业土地使用的比例从2014年的35.90％（870.70 km ²）增加到38.53％（934.50 km ^2））在2019年，森林和农业草地的土地使用面积从2014年的62.86％（1,524.01 km ²）下降到2019年的57.76％（1,400.99 km ²）。类似地，模拟结果表明，住宅和工业-商业用地将增加到到2030年和2050年分别达到42.86％（1,037.139 km ²）和44.53％（1,079.99 km ²）。同期，预计森林和草地农业用地将从53.53％（1298.53 km ²）下降到51.76％（1255.49 km ²）。空间变化将主要发生在墨尔本的北部和西北部。