Fig 1.
Methodological framework for multi-decadal LCZ classification in Sydney. The process includes (1) manual generation of training polygons, (2) LCZ classification using Landsat images and machine learning, and (3) visual and statistical validation, where OA represents the overall accuracy or the percentage of correctly classified pixels out of the total pixels. Each colored box represents a distinct step in LCZ generation. Final TAs refer to the revised training areas used to generate the final LCZ maps. In the multi-year classification, 2020 TAs are overlaid on 2015 imagery, revised/remove and added TAs to reflect changes. These updated TAs then proceed through steps (1) to (3). The final TAs for 2015 are then used to classify 2010, and so forth for earlier years. Base map and data from OpenStreetMap and OpenStreetMap Foundation.
Fig 2.
LCZ map accuracy metrics from 1990 to 2020, including overall accuracy (OA), overall accuracy for urban classes (OAu), overall accuracy for built versus natural classes (OAbu), and weighted accuracy (OAw) on the top panel across the given time periods. The bottom panel shows the distribution of training area polygons across LCZ classes for each year, with the x-axis indicating the year and the corresponding total number of training areas (TAs) for that year. The colour in the histogram (lower panel) represents the TAs from different LCZs shown in Fig 3. Here, 1 to 10 represent urban LCZ, and A to G represent natural LCZ.
Fig 3.
Multi-decadal LCZ classification of Sydney from 1990 to 2020 at five-year intervals (a-g).
The classification scheme includes ten urban classes and seven natural classes. Darker shades of red represent higher-density urban areas. The black outline delineates the Greater Sydney Region boundary, for which the shapefile was acquired from the NSW Planning Portal (https://www.planningportal.nsw.gov.au/opendata/dataset/planning-district-boundaries) [56]. Latitude and longitude are displayed in degrees south and east, respectively.
Fig 4.
LCZ map of Sydney in 2020 with close-up views highlighting significant urban changes from 1990 to 2020.
Areas of interest include (a) Parramatta, (b) North Sydney, St. Leonards, and Chatswood, and (c) Gregory Hills (southwestern Sydney). In panel (b), the blue circle covers Chatswood, St Leonards and North Sydney. The blue rectangle delineates the extent of the area of interest.
Fig 5.
Normalized changes in land cover categories for urban LCZ classes from 1990 to 2020 over Sydney.
This normalization is performed by calculating the ratio of each year’s area to the highest recorded value for that category within the period. A normalized value close to 1 indicates the highest area for a particular category in a given year.
Fig 6.
Transition matrix of urban Local Climate Zone (LCZ) classes (LCZ 1–10) in the Sydney from 1990 to 2020.
The colors represent the different LCZ classes. Each column shows a five-year transition period, while the final column presents the transition from 1990 to 2020. Each row corresponds to an individual LCZ class (LCZ 1–10). The pie charts in each cell illustrate how the land cover of a specific LCZ class changed (%) during the respective period. For example, the pie chart for compact highrise (LCZ 1) in the 1990–1995 column shows the composition of LCZ 1 in 1995, indicating the proportion that remained LCZ 1 from 1990 and the proportion that transitioned from other classes by 1995.
Fig 7.
The chord diagram represents the transition among LCZ categories in the Sydney from 1990 to 2020.
We calculated the absolute number of pixels that were LCZ X in 1990 and turned to LCZ Y in 2020. The thickness of the connecting arrow (chords) indicates the magnitude of the transitions from one class to another over time.