3
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      What you see is what you breathe? Estimating air pollution spatial variation using street level imagery

      research-article

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          High spatial resolution information on urban air pollution levels is unavailable in many areas globally, partially due to high input data needs of existing estimation approaches. Here we introduce a computer vision method to estimate annual means for air pollution levels from street level images. We used annual mean estimates of NO 2 and PM 2.5 concentrations from locally calibrated models as labels from London, New York, and Vancouver to allow for compilation of a sufficiently large dataset (~250k images for each city). Our experimental setup is designed to quantify intra and intercity transferability of image-based model estimates. Performances were high and comparable to traditional land-use regression (LUR) and dispersion models when training and testing on images from the same city (R 2 values between 0.51 and 0.95 when validated on data from ground monitoring stations). Like LUR models, transferability of models between cities in different geographies is more difficult. Specifically, transferability between the three cities i.e., London, New York, and Vancouver, which have similar pollution source profiles were moderately successful (R 2 values between zero and 0.67). Comparatively, performances when transferring models trained on these cities with very different source profiles i.e., Accra in Ghana and Hong Kong were lower (R 2 between zero and 0.21) suggesting the need for local calibration with local calibration using additional measurement data from cities that share similar source profiles.

          Related collections

          Most cited references41

          • Record: found
          • Abstract: found
          • Article: not found

          Deep learning.

          Deep learning allows computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction. These methods have dramatically improved the state-of-the-art in speech recognition, visual object recognition, object detection and many other domains such as drug discovery and genomics. Deep learning discovers intricate structure in large data sets by using the backpropagation algorithm to indicate how a machine should change its internal parameters that are used to compute the representation in each layer from the representation in the previous layer. Deep convolutional nets have brought about breakthroughs in processing images, video, speech and audio, whereas recurrent nets have shone light on sequential data such as text and speech.
            Bookmark
            • Record: found
            • Abstract: found
            • Article: found
            Is Open Access

            Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015

            Summary Background Exposure to ambient air pollution increases morbidity and mortality, and is a leading contributor to global disease burden. We explored spatial and temporal trends in mortality and burden of disease attributable to ambient air pollution from 1990 to 2015 at global, regional, and country levels. Methods We estimated global population-weighted mean concentrations of particle mass with aerodynamic diameter less than 2·5 μm (PM2·5) and ozone at an approximate 11 km × 11 km resolution with satellite-based estimates, chemical transport models, and ground-level measurements. Using integrated exposure–response functions for each cause of death, we estimated the relative risk of mortality from ischaemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, lung cancer, and lower respiratory infections from epidemiological studies using non-linear exposure–response functions spanning the global range of exposure. Findings Ambient PM2·5 was the fifth-ranking mortality risk factor in 2015. Exposure to PM2·5 caused 4·2 million (95% uncertainty interval [UI] 3·7 million to 4·8 million) deaths and 103·1 million (90·8 million 115·1 million) disability-adjusted life-years (DALYs) in 2015, representing 7·6% of total global deaths and 4·2% of global DALYs, 59% of these in east and south Asia. Deaths attributable to ambient PM2·5 increased from 3·5 million (95% UI 3·0 million to 4·0 million) in 1990 to 4·2 million (3·7 million to 4·8 million) in 2015. Exposure to ozone caused an additional 254 000 (95% UI 97 000–422 000) deaths and a loss of 4·1 million (1·6 million to 6·8 million) DALYs from chronic obstructive pulmonary disease in 2015. Interpretation Ambient air pollution contributed substantially to the global burden of disease in 2015, which increased over the past 25 years, due to population ageing, changes in non-communicable disease rates, and increasing air pollution in low-income and middle-income countries. Modest reductions in burden will occur in the most polluted countries unless PM2·5 values are decreased substantially, but there is potential for substantial health benefits from exposure reduction. Funding Bill & Melinda Gates Foundation and Health Effects Institute.
              Bookmark
              • Record: found
              • Abstract: not found
              • Article: not found

              The Lancet Commission on pollution and health

                Bookmark

                Author and article information

                Journal
                101624426
                Remote Sens (Basel)
                Remote Sens (Basel)
                Remote sensing
                2072-4292
                17 July 2022
                11 September 2023
                16 September 2023
                : 14
                : 14
                : 3429
                Affiliations
                [1 ]Imperial College London
                [2 ]ETH Zurich
                [3 ]Universities Space Research Association (USRA)
                [4 ]NASA Ames Research Center
                [5 ]University of Massachusetts Amherst
                [6 ]Arizona State University
                [7 ]San Diego State University
                [8 ]UC Berkeley
                [9 ]University of Ghana
                [10 ]University of British Columbia
                [11 ]University of Washington
                Author notes
                [* ]Correspondence: sesra@ 123456ethz.ch
                Article
                EMS187696
                10.3390/rs14143429
                7615101
                37719470
                fe9baa50-8d12-4e90-9557-12fe985f6c51

                Submitted for possible open access publication under the terms and conditions of the Creative Commons Attribution (CC BY) license ( https://creativecommons.org/licenses/by/4.0/).

                This work is licensed under a BY 4.0 International license.

                History
                Categories
                Article

                computer vision,deep learning,street images,air pollution,data science,transferability,urban pollution

                Comments

                Comment on this article