Big Remote Sensing Data and Machine Learning for Assessing 21st Century Flooding and Socioeconomic Exposures
| dc.contributor.author | Sherpa, Sonam Futi | en |
| dc.contributor.committeechair | Shirzaei, Manoochehr | en |
| dc.contributor.committeemember | Weiss, Robert | en |
| dc.contributor.committeemember | Willis, Michael J. | en |
| dc.contributor.committeemember | Werth, Susanna | en |
| dc.contributor.department | Geosciences | en |
| dc.date.accessioned | 2023-04-29T08:00:08Z | en |
| dc.date.available | 2023-04-29T08:00:08Z | en |
| dc.date.issued | 2023-04-28 | en |
| dc.description.abstract | Over the past decades, we have seen escalating costs associated with the direct socioeconomic impacts of hydrometeorological events and climate extremes such as flooding, rising sea levels due to climate change, solid earth changes, and other anthropogenic activities. With the increasing population in the era of changing climate, the number of people suffering from exposure to extreme events and sea level rise is expected to increase over the years. To develop resilience plans and mitigation strategies, hindcast exposure models, and calculate the insurance payouts, accurate maps of flooding extent and socioeconomic exposure at management-relevant resolution (102m) are needed. The growing number and continually improving coverage of Earth-observing satellites, an extensive archive of big data, and machine learning approaches have transformed the community's capacity to timely respond to flooding and water security concerns. However, in the case of flood extent mapping, most flood mapping algorithms estimate flood extent in the form of a binary map and do not provide any information on the uncertainty associated with the pixel class. Additionally, in the case of coastal inundation from sea level rise, most future projections of sea-level rise lack an accurate estimate of vertical land motion and pose a significant challenge to flood risk management plans. In this dissertation, I explore an extensive archive of available remotely sensed space-borne. synthetic aperture radar (SAR) and interferometric SAR measurements for 1) Large-scale flood extent mapping and exposure utilizing machine learning approaches and Bayesian framework to obtain probabilistic flood maps for the 2019 flood of Iran and 2018 flood of India and 2) Assessment of relative sea-level rise flooding for coastal disaster resilience in the Chesapeake Bay. Lastly, I investigate how climate change affects hydrology and cryosphere to 3) understand cryosphere-climate interaction for hazard risk and water resources management. | en |
| dc.description.abstractgeneral | Flooding increased exponentially in recent decades due to changes in climate and human activities. With an increasing number of people and flooding events, exposure to such events has been enhanced. The presence of satellites in space, the increase in revisit-time, and better tools and techniques to map flood extents have transformed society's ability to respond to hazard and water-related issues. To develop risk management plans and project how many people will be affected by hazards, and calculate the insurance payouts, accurate maps of flooding extent, and socioeconomic exposure at management-relevant resolution are needed. However, in terms of flood mapping, most flood maps do not provide information on how much water is there on a particular map. In addition, in the case of coastal flooding coming from sea level changes, current methods for future scenarios of flooding, do not accurately account for how the ocean is rising with respect to the land-encompassing movement of the land. This causes a significant challenge to coastal flood risk management plans. Therefore, in this dissertation, I explore large datasets from satellites for 1) Accurate flood extent mapping and 2) Estimation of coastal flood from relative sea level rise. Lastly, I also, examine how climate change is affecting ice and water changes to 3) Understand the role of climate on the water for hazard risk and water management. | en |
| dc.description.degree | Doctor of Philosophy | en |
| dc.format.medium | ETD | en |
| dc.identifier.other | vt_gsexam:36774 | en |
| dc.identifier.uri | http://hdl.handle.net/10919/114855 | en |
| dc.language.iso | en | en |
| dc.publisher | Virginia Tech | en |
| dc.rights | In Copyright | en |
| dc.rights.uri | http://rightsstatements.org/vocab/InC/1.0/ | en |
| dc.subject | Remote Sensing | en |
| dc.subject | Flooding | en |
| dc.subject | Machine Learning | en |
| dc.subject | Climate Sciences | en |
| dc.subject | Solid Earth | en |
| dc.title | Big Remote Sensing Data and Machine Learning for Assessing 21st Century Flooding and Socioeconomic Exposures | en |
| dc.type | Dissertation | en |
| thesis.degree.discipline | Geosciences | en |
| thesis.degree.grantor | Virginia Polytechnic Institute and State University | en |
| thesis.degree.level | doctoral | en |
| thesis.degree.name | Doctor of Philosophy | en |
Files
Original bundle
1 - 1 of 1