Chen, T.-H. K., Pandey, B., & Seto, K. C. (2023). Detecting subpixel human settlements in mountains using deep learning: A case of the Hindu Kush Himalaya 1990–2020. Remote Sensing of Environment, 294, 113625–. https://doi.org/10.1016/j.rse.2023.113625
View Publication
Abstract
The majority of future population growth in mountains will occur in small- and medium-sized cities and towns and affect vulnerable ecosystems. However, mountain settlements are often omitted from global land cover analyses due to the low spatial resolution of satellite images, which cannot resolve the small scale of mountains settlements. This study demonstrates, for the first time, the potential of deep learning to detect human settlements in mountains at the sub-pixel level, based on Landsat satellite imagery. We hypothesized that adding spatial and temporal features could improve the detection of mountain settlements since spectral information alone led to inaccurate results. For spatial features, we compared a U-shaped neural network (U-Net), a deep learning algorithm that automatically learns spatial features, with a simple random forest (RF) algorithm. Then, we assessed whether temporal features would increase accuracy by comparing two input datasets, multispectral imagery and temporal features from the Continuous Change Detection and Classification (CCDC) algorithm. We evaluated each method by calculating the accuracies of (1) the binary settlement footprint, (2) the subpixel estimates of impervious surfaces, and (3) urban growth. We tested the accuracies using visually interpreted datasets from time-series Google Earth images across the Hindu Kush Himalaya that were not used for training to evaluate model transferability. The U-Net successfully improved mountain settlement mapping compared to the random forest, with a substantial discrepancy in small settlements. The time-series results from the U-Net successfully captured long-term urban growth but fewer short-term changes. Contrary to expectations, the CCDC temporal features reduced the accuracy of mountain settlement mapping due to frequent cloud cover in hilly areas. Our subpixel analysis reveals that the built-up area of the Hindu Kush Himalaya has expanded at a rate of 61 km2 per year from 1990 to 2020, which is about twice the estimate of the Global Human Settlement Layer using binary urban/non-urban classifications.
Keywords
Urban land cover; Land cover fraction; Peri-urban; Built-up area; Subpixel mapping; Machine learning; Time-series; Himalaya; CCDC
The Urban Infrastructure Lab researchers have released a report on a Cascadia region high-speed rail project. College of Built Environments faculty Jan Whittington and Qing Shen were authors on the report, along with 3 Interdisciplinary Ph.D. in Urban Design and Planning students (Siman Ning, Haoyu Yue, and Chin-Wei Chen), and a Master of Urban Planning candidate (Richard McMichael). This report examines the successes and lessons learned from existing high-speed rail projects in Europe and Asia, including 50 hours of interviews…
Chen, T.-H. K., Horsdal, H. T., Samuelsson, K., Closter, A. M., Davies, M., Barthel, S., Pedersen, C. B., Prishchepov, A. V., & Sabel, C. E. (2023). Higher depression risks in medium- than in high-density urban form across Denmark. Science Advances, 9(21), eadf3760–eadf3760. https://doi.org/10.1126/sciadv.adf3760
View Publication
Abstract
Urban areas are associated with higher depression risks than rural areas. However, less is known about how different types of urban environments relate to depression risk. Here, we use satellite imagery and machine learning to quantify three-dimensional (3D) urban form (i.e., building density and height) over time. Combining satellite-derived urban form data and individual-level residential addresses, health, and socioeconomic registers, we conduct a case-control study (n = 75,650 cases and 756,500 controls) to examine the association between 3D urban form and depression in the Danish population. We find that living in dense inner-city areas did not carry the highest depression risks. Rather, after adjusting for socioeconomic factors, the highest risk was among sprawling suburbs, and the lowest was among multistory buildings with open space in the vicinity. The finding suggests that spatial land-use planning should prioritize securing access to open space in densely built areas to mitigate depression risks.
UW Department of Urban Design and Planning in the College of Built Environments.
The Population Health Initiative announced the summer 2023 cohort of applied research fellows. Among the group of 5 students (3 graduate and 2 undergraduate) is Pamela Lim from the College of Built Environments Urban Design & Planning department. These students will spend 10 weeks over the summer working collaboratively with King County’s demographer and Public Health – Seattle & King County’s Assessment, Policy Development and Evaluation Unit with the support of the Population Health Applied Research Fellowship. The team will…
EarthLab selected the 2023-2024 Innovation Grant Awardees in April 2023. One of the projects chosen includes College of Built Environments researchers on the interdisciplinary team. The project description and research team is detailed below. “Cultivating Transdisciplinary Support for Equitable and Resilient Floodplain Solutions” Project Description: In 2021 a massive flood on the Nooksack River left a trail of destruction in its wake. Floods are the most expensive natural hazard in Washington State, a risk that is exacerbated by climate change….
The Population Health Initiative announced 12 climate change planning grant awardees. Of those 12 teams, 4 include College of Built Environments researchers. Descriptions of their projects are below. Read the CBE News story here. Linking Climate Adaptation and Public Health Outcomes in Yavatmal, Maharashtra Investigators Sameer H. Shah, Environmental and Forest Sciences Celina Balderas Guzmán, Landscape Architecture Pronoy Rai, Portland State University Project abstract This proposal collects primary interview data with landed and landless agriculturalists in Yavatmal district in…
College of Built Environments researchers are selected for inaugural cohort of the Urban@UW Research to Action Collaboratory (RAC). Throughout the next 18 months, Urban@UW will work with these teams and provide seed funds, dedicated time to build team cohesion and collaboration skills, and foster opportunities for peer support and shared resources and learning. The project below was one of two projects selected for this cohort. See the full story here. Just Circular Communities: A Resiliency Framework to Support a Just…
Fan, R., Yocom, K. P., & Guo, Y. (2023). Utilizing Fractal Dimensions as Indicators to Detect Elements of Visual Attraction: A Case Study of the Greenway along Lake Taihu, China. Land (Basel), 12(4), 883–. https://doi.org/10.3390/land12040883
View Publication
Abstract
It is widely acknowledged that the quality of greenway landscape resources enhances the visual appeal of people. While most studies have evaluated visual perception and preference, few have considered the relationship between the distribution of greenways in relation to the proximity of water bodies such as lakes and rivers. Such an investigation requires an in-depth analysis of how to plan and design greenways in order to better enhance people's willingness to access and utilize them. In this research we propose specific color brightness and contour visual attraction elements to further discuss the quality of greenway landscape resources in the rapidly urbanizing Lake Taihu region of China. Specifically, we utilize a common method in fractal theory analysis called counting box dimension to calculate and analyze the sample images. The method generates data on fractal dimension (FD) values of two elements; the optimal fractal dimension threshold range; the characteristics exhibited by the maximum and minimum fractal dimension values in the greenway landscape; and the relationship between the two visual attraction elements allowing us to derive distribution of the greenway and water bodies. The results reveal that greenway segments with high values of the visual attraction element of color brightness fractal dimension (FD) are significantly closer to the lake than those subject to high values of the visual attraction element. Some segments are even close to the lake surface, which is because the glare from the direct sunlight and the reflection from the lake surface superimposed on each other, so that the greenway near the lake surface is also affected by the brightness and shows the result of high color brightness values. However, the greenway segments with high values of contour element FD are clearly more influenced by plants and other landscape elements. This is due to the rich self-similarity of the plants themselves. Most of the greenway segments dominated by contour elements are distant from the lake surface. Both color brightness and contour elements are important indicators of the quality of the visual resources of the Lake Taihu Greenway landscape. This reveals that the determination of the sub-dimensional values of color brightness (1.7608, 1.9337) and contour (1.7230, 1.9006) visual attraction elements and the optimal threshold range (1.7608, 1.9006) can provide theoretical implications for the landscape planning and design of lake-ring type greenways and practical implications for assessing the quality of visual resources in greenway landscapes.
Keywords
color brightness; contour; visual attraction; fractal dimension (FD); boxplot; Lake Taihu
The Population Health Initiative announced the award of nine Tier 1 pilot grants to interdisciplinary research teams representing 10 of the University of Washington’s schools and colleges. The total award value of these grants is nearly $210,000, which includes school, department and unit matching funds. Read more in the CBE Story here. “We were extremely pleased with the range of challenges these awards will work to address,” said Ali H. Mokdad, the UW’s chief strategy officer for population health and professor of…