[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-128612-en":3,"doc-seo-128612-105":31,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},128612,962084925502,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_6f874abed73319feea01a86fa6f0fab8",8,"Research & Report","SPATIOTEMPORAL GAP-FILLING OF NASA DEEP BLUE AEROSOL OPTICAL DEPTH OVER CONUS USING THE UNET 3+ ARCHITECTURE - A THESIS","The thesis develops a spatiotemporal gap-filling approach to estimate NASA Deep Blue (DB) aerosol optical depth (AOD) over the CONUS region using a UNet 3+ architecture. Training and evaluation rely on multiple data sources, including satellite DB retrievals and auxiliary reanalysis information, with carefully defined temporal resolution, input variables, and target weighting. Model outputs are assessed via spatial statistics and scatter comparisons against AERONET measurements at 550 nm, quantifying agreement and improving coverage when DB retrievals are missing. Results also compare performance across alternative gap-filling models.","UNIVERSITY OF OKLAHOMA  \nGRADUATE COLLEGE  \nSPATIOTEMPORAL GAP-FILLING OF NASA DEEP BLUE AEROSOL  \nOPTICAL DEPTH  \nOVER CONUS USING THE UNET 3+ ARCHITECTURE  \nA THESIS  \nSUBMITTED TO THE GRADUATE FACULTY  \nin partial fulfillment of the requirements for the  \nDegree of  \nMASTER OF SCIENCE  \nBy  \nJeffrey Lee  \nNorman, Oklahoma  \nSPATIOTEMPORAL GAP-FILLING OF NASA DEEP BLUE AEROSOL  \nOPTICAL DEPTH  \nOVER CONUS USING THE UNET 3+ ARCHITECTURE  \nA THESIS APPROVED FOR THE  \nSCHOOL OF METEOROLOGY  \nBY THE COMMITTEE CONSISTING OF  \nDr. Marcela Loría-Salazar, Chair  \nDr. Amy McGovern  \nDr. Feng Xu  \nDr. Jason Furtado  \n© Copyright by Jeffrey Lee 2024 All Rights Reserved.  \nAcknowledgements  \nFirstly, I would like to thank my research advisor Dr. Marcela Loría-Salazar, who has offered me countless hours of instruction, research guidance, and mental support. This research would never have materialized without her steadfast presence.  \nI would also like to thank the members of the committee, Dr. Amy McGovern, Dr. Feng Xu, and Dr. Jason Furtado, for graciously agreeing to review my work, provide their valuable insights, and accommodate the various delays that have taken place throughout the process.  \nI would also like to thank Dr. Heather Holmes at the University of Utah and her research group for allowing me access to the CHPC supercomputing resources. Without these resources, training a model ofour size would have been an impracticality.  \nI want to thank the faculty and staff at the University of Oklahoma’s School of Meteorology for providing such a wonderful learning environment. I have learned so much about things I had never dreamed of before from the passionate and experienced educators here at SoM. The wonderful staff have also made sure my logistical, administrative, and financial issues were kept at a minimum.  \nFinally, I would like to thank my friends and family who have stuck with me through the years. You guys are the best.  \nList of Tables  \nTable 1: List of all data sources used in the training and evaluation of the UNet 3+ gap-filler... 16  \nTable 2: Temporal resolution of various input data...................................................................... 30  \nTable 3: Input variables, native resolutions, and lags provided to the model............................... 39  \nTable 4: Target AOD weights....................................................................................................... 40  \nTable 5: Statistics of MERRA-2 AOD, DB AOD, UNet 3+ AOD, and gap-filled AOD against AERONET AOD .......................................................................................................................... 46  \nTable 6: A collection of AOD gap-filling models and their performance.................................... 61  \nList of Figures  \nFigure 1: The size of aerosols, or particulate matter (PM), that can penetrate the human body (Kim et al., 2015) ............................................................................................................................ 2  \nFigure 2: VIIRS visible imagery over the Pacific Northwest on September 5, 2017, showing heavy wildfire smoke. Source: NASA Worldview......................................................................... 3  \nFigure 3: An image over northern Africa showing the presence of dust aerosols and the effectiveness of the Deep Blue wavelength at isolating these aerosols. Source: MODIS Terra.. 10  \nFigure 4: Depiction of the orbit of a polar-orbiting satellite. Source: Space Foundation............. 18  \nFigure 5: A map of the NAM domain (solid) as well as the parent domain (dashed) . We used  \ndata gridded in the solid region. Source: (“North American Mesoscale Forecast System,” 2020) ........................................................................................................................................................ 20  \nFigure 6: A sample map of the HMS smoke product for November 8, 2018, showing a large smoke plume from ","cbCaiiDfqHfdUVp2","https://ap.wps.com/l/cbCaiiDfqHfdUVp2","pdf",6271234,3,1,109,"English","en",105,"# Acknowledgements\n# List of Tables\n# List of Figures\n# Tables\n## Table 1: Data sources for UNet 3+ gap-filler training and evaluation\n## Table 2: Temporal resolution of various input data\n## Table 3: Model input variables, resolutions, and lags\n## Table 4: Target AOD weights\n## Table 5: Statistics vs. AERONET AOD\n## Table 6: Performance of gap-filling models\n# Figures\n## Figure 1: Aerosol particle size and human-body penetration\n## Figure 2: VIIRS visible imagery of wildfire smoke (Sept 5, 2017)\n## Figure 3: Dust aerosols over northern Africa and Deep Blue wavelength separation\n## Figure 4: Orbit depiction of a polar-orbiting satellite\n## Figure 5: NAM domain vs. parent domain grid usage\n## Figure 6: HMS smoke product example (Nov 8, 2018)\n## Figure 7: Proposed grid and NAM 12 km grid projections\n## Figure 8: NAM data buffers overlaying a grid cell\n## Figure 9: Histogram of log AOD values from satellite and MERRA-2\n## Figure 10: UNet 3+ predictions, DB retrievals, gap-filled AOD, and MERRA-2 AOD example\n## Figure 11: Seasonal distribution of successful DB AOD retrieval days (2012–2022)\n## Figure 12: DB AOD retrievals with supplemental MERRA-2 AOD\n## Figure 13: Selective backpropagation effect on an MLP model\n## Figure 14: Scatter plots vs. AERONET AOD (550 nm) for multiple products","[{\"question\":\"What problem does the thesis address regarding NASA Deep Blue aerosol optical depth?\",\"answer\":\"It addresses missing DB aerosol optical depth retrievals by creating a spatiotemporal gap-filling method to estimate AOD over the CONUS region.\"},{\"question\":\"What model architecture is used for the gap-filling method?\",\"answer\":\"The method uses a UNet 3+ architecture trained with satellite DB AOD and auxiliary data inputs.\"},{\"question\":\"How is the gap-filled AOD evaluated in the thesis?\",\"answer\":\"Evaluation is performed using spatial statistical analysis and scatter comparisons against AERONET AOD at 550 nm, including agreement metrics for UNet 3+ and baseline/alternative approaches.\"}]","SPATIOTEMPORAL GAP-FILLING OF NASA DEEP BLUE AEROSOL OPTICAL DEPTH OVER CONUS USING THE UNET 3+ ARCHITECTURE - A THESIS | PDF",1786002098,275,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":29},"spatiotemporal-gap-filling-of-nasa-deep-blue-aerosol-optical-depth-over-conus-using-the-unet-3-architecture-a-thesis","",{"@graph":37,"@context":86},[38,54,69],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,51],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":20},"https://docshare.wps.com/document/research-report/",{"item":52,"name":13,"@type":44,"position":53},"https://docshare.wps.com/document/spatiotemporal-gap-filling-of-nasa-deep-blue-aerosol-optical-depth-over-conus-using-the-unet-3-architecture-a-thesis/128612/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":42,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-25","2026-08-06",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What problem does the thesis address regarding NASA Deep Blue aerosol optical depth?","Question",{"text":76,"@type":77},"It addresses missing DB aerosol optical depth retrievals by creating a spatiotemporal gap-filling method to estimate AOD over the CONUS region.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What model architecture is used for the gap-filling method?",{"text":81,"@type":77},"The method uses a UNet 3+ architecture trained with satellite DB AOD and auxiliary data inputs.",{"name":83,"@type":74,"acceptedAnswer":84},"How is the gap-filled AOD evaluated in the thesis?",{"text":85,"@type":77},"Evaluation is performed using spatial statistical analysis and scatter comparisons against AERONET AOD at 550 nm, including agreement metrics for UNet 3+ and baseline/alternative approaches.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":47,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]