Based on open-source satellite images captured during Hurricane Ida, Rebecca Composto developed a mapping tool<\/a> that could help identify potentially flood-prone areas in urban settings \u2014 and empower officials with the information needed to make better decisions about where to allocate flood response and prevention resources.<\/p>\n\n\n\n Composto says that compared to rural communities, urban areas present unique challenges in collecting satellite data and tracking water flow. <\/p>\n\n\n\n For one, flooding in urban areas tends to start and end quickly \u2014 sometimes too quickly for satellites to gather enough usable data.<\/p>\n\n\n\n Meanwhile, a sea of skyscrapers, streets and sidewalks doesn\u2019t make things any easier.<\/p>\n\n\n\n \u201cTaller buildings create more shadows, which means that the satellite imagery appears darker and carries less information,\u201d says Composto, who\u2019s working toward a Ph.D. in geospatial analytics. \u201cUrban areas also have more complex hydrology, as the existence of so many drainage systems along with concrete surfaces that don\u2019t soak up water means that it\u2019s harder to predict where water accumulates.\u201d<\/p>\n\n\n\n With Mirela G. Tulbure, a Center for Geospatial Analytics Faculty Fellow and professor in NC State\u2019s College of Natural Resources, postdoctoral researcher J\u00falio Caineta, graduate research assistant Varun Tiwari and fellow geospatial analytics Ph.D. candidate Mollie D. Gaines, Composto recently published a paper<\/a>, \u201cQuantifying urban flood extent using satellite imagery and machine learning,\u201d in the journal Natural Hazards<\/em>.<\/p>\n\n\n\n Future research could focus on making the model easier to use, Composto says. She also plans to integrate a new map displaying flood depth \u2014 and make her code open source so it\u2019s easier to share with emergency-response leaders.<\/p>\n\n\n\n This article was originally published<\/a> in NC State News.<\/em> A news release<\/a> was also published in NC State News<\/em>.<\/p>\n","protected":false,"raw":"\n\n\n\n\n If it seems like certain places suffer from worse flooding these days, it\u2019s because they do. <\/p>\n\n\n\n Natural disasters have always tended to hit certain communities harder \u2014 often those already at a disadvantage, by design \u2014 and climate change has only intensified the problem, especially for big cities. Several aspects of urban areas make it more difficult to map where flooding occurs. So, in comparison to rural areas, urban areas have historically been understudied.<\/p>\n\n\n\n Modern-day emergency planners need tools to more accurately predict \u2014 and better prepare for \u2014 flooding in urban areas. A graduate student at NC State University\u2019s Center for Geospatial Analytics<\/a> has trained a machine-learning model to create maps that could someday help them do just that.<\/p>\n\n\n\n Based on open-source satellite images captured during Hurricane Ida, Rebecca Composto developed a mapping tool<\/a> that could help identify potentially flood-prone areas in urban settings \u2014 and empower officials with the information needed to make better decisions about where to allocate flood response and prevention resources.<\/p>\n\n\n\n Composto says that compared to rural communities, urban areas present unique challenges in collecting satellite data and tracking water flow. <\/p>\n\n\n\n For one, flooding in urban areas tends to start and end quickly \u2014 sometimes too quickly for satellites to gather enough usable data.<\/p>\n\n\n\n Meanwhile, a sea of skyscrapers, streets and sidewalks doesn\u2019t make things any easier.<\/p>\n\n\n\n \u201cTaller buildings create more shadows, which means that the satellite imagery appears darker and carries less information,\u201d says Composto, who\u2019s working toward a Ph.D. in geospatial analytics. \u201cUrban areas also have more complex hydrology, as the existence of so many drainage systems along with concrete surfaces that don\u2019t soak up water means that it\u2019s harder to predict where water accumulates.\u201d<\/p>\n\n\n\n With Mirela G. Tulbure, a Center for Geospatial Analytics Faculty Fellow and professor in NC State\u2019s College of Natural Resources, postdoctoral researcher J\u00falio Caineta, graduate research assistant Varun Tiwari and fellow geospatial analytics Ph.D. candidate Mollie D. Gaines, Composto recently published a paper<\/a>, \u201cQuantifying urban flood extent using satellite imagery and machine learning,\u201d in the journal Natural Hazards<\/em>.<\/p>\n\n\n\n Future research could focus on making the model easier to use, Composto says. She also plans to integrate a new map displaying flood depth \u2014 and make her code open source so it\u2019s easier to share with emergency-response leaders.<\/p>\n\n\n\n This article was originally published<\/a> in NC State News.<\/em> A news release<\/a> was also published in NC State News<\/em>.<\/p>\n"},"excerpt":{"rendered":" A Ph.D. student at the Center for Geospatial Analytics trained a machine-learning model to create maps that could help identify potentially flood-prone areas in urban settings.<\/p>\n","protected":false},"author":152,"featured_media":22408,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"source":"","ncst_custom_author":"","ncst_show_custom_author":false,"ncst_dynamicHeaderBlockName":"ncst\/default-post-header","ncst_dynamicHeaderData":"{\"caption\":\"A flooded street in Philadelphia, Pennsylvania, in the aftermath of Hurricane Ida, which caused significant damage across the northeastern U.S. in 2021.\",\"displayCategoryID\":53,\"showAuthor\":false,\"showDate\":true,\"showFeaturedVideo\":false}","ncst_content_audit_freq":"","ncst_content_audit_date":"","ncst_content_audit_display":false,"ncst_backToTopFlag":"","footnotes":"","_links_to":"","_links_to_target":""},"categories":[53,7,48,8,13,10,6],"tags":[],"class_list":["post-22407","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-creating-near-real-time-decision-analytics","category-faculty-and-staff","category-geospatial-analytics-phd","category-new-publications","category-new-research","category-spotlight","category-student"],"displayCategory":{"term_id":53,"name":"Creating Near Real-Time Decision Analytics","slug":"creating-near-real-time-decision-analytics","term_group":0,"term_taxonomy_id":53,"taxonomy":"category","description":"","parent":0,"count":41,"filter":"raw"},"acf":{"ncst_posts_meta_modified_date":null},"_links":{"self":[{"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/posts\/22407","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/users\/152"}],"replies":[{"embeddable":true,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/comments?post=22407"}],"version-history":[{"count":4,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/posts\/22407\/revisions"}],"predecessor-version":[{"id":22570,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/posts\/22407\/revisions\/22570"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/media\/22408"}],"wp:attachment":[{"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/media?parent=22407"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/categories?post=22407"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cnr.ncsu.edu\/geospatial\/wp-json\/wp\/v2\/tags?post=22407"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}![\"A](\"https:\/\/news.ncsu.edu\/wp-content\/uploads\/2024\/08\/map_webpage-1024x576.png\")
<\/a><\/a>