{"id":{"repo_id":"arizona-thes","oai_identifier":"oai:repository.arizona.edu:10150/668077"},"canonical_url":"https://search.dev.ndltd.org/etd/arizona-thes/oai:repository.arizona.edu:10150/668077","repository":{"repo_id":"arizona-thes","name":"University of Arizona","base_url":"https://repository.arizona.edu/oai/request"},"display":{"title":"Improving Global Satellite Precipitation Products Utilizing Machine Learning","abstract":"This dissertation investigates applications of machine learning and deep learning to improve global satellite precipitation products. This includes providing practical guidance and analysis to determine which sensor or algorithm can provide the most effective precipitation input to the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM; IMERG; or other global precipitation products such as the Global Precipitation Climatology Project; GPCP) in high latitudes (i.e., poleward of 60°S/N). The possibility of mitigating the latency gap of satellite precipitation products through deep-learning-based nowcasting is also investigated in this dissertation.IMERG is the most popular product among GPM products and is used for various scientific analyses/applications. The latest version of the IMERG (V06) was extended to the poles but still has gaps over snow and ice surfaces poleward of 60°S/N, where it also shows significant underestimation. This is mainly due to the low-quality level-2 precipitation retrievals used in IMERG (currently only passive microwave [PMW] sensors are used in high latitudes). While infrared data from geostationary satellites are used to fill some of the gaps within 60°S/N, outside of this range no infrared estimates have been used in IMERG, mainly because high-quality geostationary observation and precipitation estimates from that are not available and using single polar-orbital infrared satellites (e.g., Atmospheric Infrared Sensor [AIRS] already used in GPCP) cannot provide temporal sampling required by IMERG. So far, no study has carefully assessed/ranked available infrared products (or alternative retrievals) in high latitudes to provide practical guidance and recommendation to the IMERG and GPCP teams for enhancing their products in high latitudes. In this dissertation, I performed a comprehensive analysis to collect and assess available precipitation estimates from the AIRS and the recently developed precipitation estimates from AVHRR using the CloudSat precipitation product and the most effective machine-learning methods. In addition, I performed a comprehensive intercomparison among the PMW and infrared precipitation products to determine their skill in capturing high-latitude precipitation and provided a regional and seasonal ranking of the estimates presented in Appendix A. Regarding the latency gap of satellite precipitation products, specifically for IMERG, there is usually a few hours gap between when satellite observations are made and when precipitation estimates become available through IMERG. This is mainly due to the time required to process the observations and run retrieval algorithms. For specific applications such as flood warning systems, timely precipitation estimates are needed. To overcome this issue, we developed deep-learning-based nowcasting systems (hereafter NowCasting-nets) that can forecast precipitation based on the latest IMERG observations. This research is presented in Appendix B of this dissertation. The research conducted here directly contributes to the improvement of the NASA GPM IMERG and NASA GPCP products. These products are important not only for the present weather observations, but also for long-term climate analysis, either by themselves or through their contribution to the climate products. The effort will also utilize diverse but complementary measurements (mainly invested by NASA) to assess precipitation estimates over high-latitude regions as part of its evaluation approach.","abstract_html":"This dissertation investigates applications of machine learning and deep learning to improve global satellite precipitation products. This includes providing practical guidance and analysis to determine which sensor or algorithm can provide the most effective precipitation input to the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM; IMERG; or other global precipitation products such as the Global Precipitation Climatology Project; GPCP) in high latitudes (i.e., poleward of 60°S/N). The possibility of mitigating the latency gap of satellite precipitation products through deep-learning-based nowcasting is also investigated in this dissertation.IMERG is the most popular product among GPM products and is used for various scientific analyses/applications. The latest version of the IMERG (V06) was extended to the poles but still has gaps over snow and ice surfaces poleward of 60°S/N, where it also shows significant underestimation. This is mainly due to the low-quality level-2 precipitation retrievals used in IMERG (currently only passive microwave [PMW] sensors are used in high latitudes). While infrared data from geostationary satellites are used to fill some of the gaps within 60°S/N, outside of this range no infrared estimates have been used in IMERG, mainly because high-quality geostationary observation and precipitation estimates from that are not available and using single polar-orbital infrared satellites (e.g., Atmospheric Infrared Sensor [AIRS] already used in GPCP) cannot provide temporal sampling required by IMERG. So far, no study has carefully assessed/ranked available infrared products (or alternative retrievals) in high latitudes to provide practical guidance and recommendation to the IMERG and GPCP teams for enhancing their products in high latitudes. In this dissertation, I performed a comprehensive analysis to collect and assess available precipitation estimates from the AIRS and the recently developed precipitation estimates from AVHRR using the CloudSat precipitation product and the most effective machine-learning methods. In addition, I performed a comprehensive intercomparison among the PMW and infrared precipitation products to determine their skill in capturing high-latitude precipitation and provided a regional and seasonal ranking of the estimates presented in Appendix A. Regarding the latency gap of satellite precipitation products, specifically for IMERG, there is usually a few hours gap between when satellite observations are made and when precipitation estimates become available through IMERG. This is mainly due to the time required to process the observations and run retrieval algorithms. For specific applications such as flood warning systems, timely precipitation estimates are needed. To overcome this issue, we developed deep-learning-based nowcasting systems (hereafter NowCasting-nets) that can forecast precipitation based on the latest IMERG observations. This research is presented in Appendix B of this dissertation. The research conducted here directly contributes to the improvement of the NASA GPM IMERG and NASA GPCP products. These products are important not only for the present weather observations, but also for long-term climate analysis, either by themselves or through their contribution to the climate products. The effort will also utilize diverse but complementary measurements (mainly invested by NASA) to assess precipitation estimates over high-latitude regions as part of its evaluation approach.","abstract_has_math":false,"creators":["Ehsani, Mohammad Reza"],"institution":"The University of Arizona.","degree_name":"Ph.D.","degree_level":"doctoral","degree_discipline":"Graduate College","degree_department":null,"school":null,"contributors":[],"advisors":["Behrangi, Ali"],"committee_chairs":[],"committee_members":["Gupta, Hoshin","Huffman, George","Bethard, Steven","Niu, Guo-Yue"],"year":2023,"date_issued":"2023","date_published":"2023","updated_at":"2026-07-24T00:57:45Z","subjects":["Climate","Deep Learning","IMERG","Machine Learning","NASA GPM","Precipitation"],"languages":["en"],"rights":["Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author."],"rights_urls":["http://rightsstatements.org/vocab/InC/1.0/"],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/10150/668077","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Behrangi, Ali"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Gupta, Hoshin","Huffman, George","Bethard, Steven","Niu, Guo-Yue"]},{"key":"dc:creator","label":"Author","values":["Ehsani, Mohammad Reza"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2023-05-10T00:23:21Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2023-05-10T00:23:21Z"]},{"key":"dc:date.issued","label":"Date","values":["2023"]},{"key":"dc:publisher","label":"Institution","values":["The University of Arizona."]},{"key":"dc:type","label":"Dc Type","values":["text","Electronic Dissertation"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Graduate College","Hydrology"]},{"key":"thesis:degree_level","label":"Degree Level","values":["doctoral"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Arizona"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Climate","Deep Learning","IMERG","Machine Learning","NASA GPM","Precipitation"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author."]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/vocab/InC/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["http://hdl.handle.net/10150/668077"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["This dissertation investigates applications of machine learning and deep learning to improve global satellite precipitation products. This includes providing practical guidance and analysis to determine which sensor or algorithm can provide the most effective precipitation input to the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM; IMERG; or other global precipitation products such as the Global Precipitation Climatology Project; GPCP) in high latitudes (i.e., poleward of 60°S/N). The possibility of mitigating the latency gap of satellite precipitation products through deep-learning-based nowcasting is also investigated in this dissertation.IMERG is the most popular product among GPM products and is used for various scientific analyses/applications. The latest version of the IMERG (V06) was extended to the poles but still has gaps over snow and ice surfaces poleward of 60°S/N, where it also shows significant underestimation. This is mainly due to the low-quality level-2 precipitation retrievals used in IMERG (currently only passive microwave [PMW] sensors are used in high latitudes). While infrared data from geostationary satellites are used to fill some of the gaps within 60°S/N, outside of this range no infrared estimates have been used in IMERG, mainly because high-quality geostationary observation and precipitation estimates from that are not available and using single polar-orbital infrared satellites (e.g., Atmospheric Infrared Sensor [AIRS] already used in GPCP) cannot provide temporal sampling required by IMERG. So far, no study has carefully assessed/ranked available infrared products (or alternative retrievals) in high latitudes to provide practical guidance and recommendation to the IMERG and GPCP teams for enhancing their products in high latitudes. In this dissertation, I performed a comprehensive analysis to collect and assess available precipitation estimates from the AIRS and the recently developed precipitation estimates from AVHRR using the CloudSat precipitation product and the most effective machine-learning methods. In addition, I performed a comprehensive intercomparison among the PMW and infrared precipitation products to determine their skill in capturing high-latitude precipitation and provided a regional and seasonal ranking of the estimates presented in Appendix A. Regarding the latency gap of satellite precipitation products, specifically for IMERG, there is usually a few hours gap between when satellite observations are made and when precipitation estimates become available through IMERG. This is mainly due to the time required to process the observations and run retrieval algorithms. For specific applications such as flood warning systems, timely precipitation estimates are needed. To overcome this issue, we developed deep-learning-based nowcasting systems (hereafter NowCasting-nets) that can forecast precipitation based on the latest IMERG observations. This research is presented in Appendix B of this dissertation. The research conducted here directly contributes to the improvement of the NASA GPM IMERG and NASA GPCP products. These products are important not only for the present weather observations, but also for long-term climate analysis, either by themselves or through their contribution to the climate products. The effort will also utilize diverse but complementary measurements (mainly invested by NASA) to assess precipitation estimates over high-latitude regions as part of its evaluation approach."]},{"key":"dc:title","label":"Title","values":["Improving Global Satellite Precipitation Products Utilizing Machine Learning"]}]}],"canonical_facts":{"dc:contributor.advisor":["Behrangi, Ali"],"dc:contributor.committeemember":["Gupta, Hoshin","Huffman, George","Bethard, Steven","Niu, Guo-Yue"],"dc:creator":["Ehsani, Mohammad Reza"],"dc:date.accessioned":["2023-05-10T00:23:21Z"],"dc:date.available":["2023-05-10T00:23:21Z"],"dc:date.issued":["2023"],"dc:description.abstract":["This dissertation investigates applications of machine learning and deep learning to improve global satellite precipitation products. This includes providing practical guidance and analysis to determine which sensor or algorithm can provide the most effective precipitation input to the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM; IMERG; or other global precipitation products such as the Global Precipitation Climatology Project; GPCP) in high latitudes (i.e., poleward of 60°S/N). The possibility of mitigating the latency gap of satellite precipitation products through deep-learning-based nowcasting is also investigated in this dissertation.IMERG is the most popular product among GPM products and is used for various scientific analyses/applications. The latest version of the IMERG (V06) was extended to the poles but still has gaps over snow and ice surfaces poleward of 60°S/N, where it also shows significant underestimation. This is mainly due to the low-quality level-2 precipitation retrievals used in IMERG (currently only passive microwave [PMW] sensors are used in high latitudes). While infrared data from geostationary satellites are used to fill some of the gaps within 60°S/N, outside of this range no infrared estimates have been used in IMERG, mainly because high-quality geostationary observation and precipitation estimates from that are not available and using single polar-orbital infrared satellites (e.g., Atmospheric Infrared Sensor [AIRS] already used in GPCP) cannot provide temporal sampling required by IMERG. So far, no study has carefully assessed/ranked available infrared products (or alternative retrievals) in high latitudes to provide practical guidance and recommendation to the IMERG and GPCP teams for enhancing their products in high latitudes. In this dissertation, I performed a comprehensive analysis to collect and assess available precipitation estimates from the AIRS and the recently developed precipitation estimates from AVHRR using the CloudSat precipitation product and the most effective machine-learning methods. In addition, I performed a comprehensive intercomparison among the PMW and infrared precipitation products to determine their skill in capturing high-latitude precipitation and provided a regional and seasonal ranking of the estimates presented in Appendix A. Regarding the latency gap of satellite precipitation products, specifically for IMERG, there is usually a few hours gap between when satellite observations are made and when precipitation estimates become available through IMERG. This is mainly due to the time required to process the observations and run retrieval algorithms. For specific applications such as flood warning systems, timely precipitation estimates are needed. To overcome this issue, we developed deep-learning-based nowcasting systems (hereafter NowCasting-nets) that can forecast precipitation based on the latest IMERG observations. This research is presented in Appendix B of this dissertation. The research conducted here directly contributes to the improvement of the NASA GPM IMERG and NASA GPCP products. These products are important not only for the present weather observations, but also for long-term climate analysis, either by themselves or through their contribution to the climate products. The effort will also utilize diverse but complementary measurements (mainly invested by NASA) to assess precipitation estimates over high-latitude regions as part of its evaluation approach."],"dc:identifier.uri":["http://hdl.handle.net/10150/668077"],"dc:language.iso":["en"],"dc:publisher":["The University of Arizona."],"dc:rights":["Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author."],"dc:rights.uri":["http://rightsstatements.org/vocab/InC/1.0/"],"dc:subject":["Climate","Deep Learning","IMERG","Machine Learning","NASA GPM","Precipitation"],"dc:title":["Improving Global Satellite Precipitation Products Utilizing Machine Learning"],"dc:type":["text","Electronic Dissertation"],"thesis:degree_discipline":["Graduate College","Hydrology"],"thesis:degree_level":["doctoral"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Arizona"]},"updated_at":"2026-07-24T00:57:45Z"}