{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/90603"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/90603","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Detection and characterization of forest disturbances in California","abstract":"Natural and anthropogenic forest disturbances are a major influence on the global carbon cycle, with the local and global impacts of a disturbance event depending in large part on the timing, intensity, and cause of the disturbance. With some disturbance agents expected to increase in frequency and severity under the influence of climate change, knowledge of disturbance events and trends has become particularly crucial for a variety of scientific, political and management-related needs. Many studies have made use of time series analyses of Landsat imagery to study forest disturbance events at a variety of scales, but few have endeavored to attribute specific causal information to the disturbances detected, particularly at a subannual temporal scale. The purpose of this research is to investigate the suitability of a Landsat time series approach for detecting and describing the causes of disturbance events across the heterogeneous, forested landscapes of California. We used a random forest machine learning algorithm to relate normalized Landsat time series from 29,909 Landsat scenes extracted at the locations of known disturbance events with the cause of each event, including classes of logging, fire, pest damage, land use conversion, and no change. Validation of the model’s thematic accuracy performed using an independent dataset demonstrated an overall accuracy of 89.2% and a kappa value of 0.77, with by-class user’s accuracies ranging from 65% to 92% and by-class producer's accuracies ranging from 37% to 98%. We then applied this model to a sample footprint in California using all Landsat imagery available, generating a pixel-by-pixel disturbance record for every pixel from 1990 to 2010. While the model demonstrates some limitations in efficiency and temporal accuracy, it also highlights the capacity for disturbance detection at a subannual interval to offer insight into forest disturbance dynamics at a finer temporal scale.","abstract_html":"Natural and anthropogenic forest disturbances are a major influence on the global carbon cycle, with the local and global impacts of a disturbance event depending in large part on the timing, intensity, and cause of the disturbance. With some disturbance agents expected to increase in frequency and severity under the influence of climate change, knowledge of disturbance events and trends has become particularly crucial for a variety of scientific, political and management-related needs. Many studies have made use of time series analyses of Landsat imagery to study forest disturbance events at a variety of scales, but few have endeavored to attribute specific causal information to the disturbances detected, particularly at a subannual temporal scale. The purpose of this research is to investigate the suitability of a Landsat time series approach for detecting and describing the causes of disturbance events across the heterogeneous, forested landscapes of California. We used a random forest machine learning algorithm to relate normalized Landsat time series from 29,909 Landsat scenes extracted at the locations of known disturbance events with the cause of each event, including classes of logging, fire, pest damage, land use conversion, and no change. Validation of the model’s thematic accuracy performed using an independent dataset demonstrated an overall accuracy of 89.2% and a kappa value of 0.77, with by-class user’s accuracies ranging from 65% to 92% and by-class producer's accuracies ranging from 37% to 98%. We then applied this model to a sample footprint in California using all Landsat imagery available, generating a pixel-by-pixel disturbance record for every pixel from 1990 to 2010. While the model demonstrates some limitations in efficiency and temporal accuracy, it also highlights the capacity for disturbance detection at a subannual interval to offer insight into forest disturbance dynamics at a finer temporal scale.","abstract_has_math":false,"creators":["Reents, Courtney E"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Geography","degree_department":null,"school":null,"contributors":["Greenberg, Jonathan A."],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2016,"date_issued":"2016-07-07T19:54:31Z","date_published":"2016-07-07T19:54:31Z","updated_at":"2026-07-22T22:26:34Z","subjects":["forest disturbance","change detection","multitemporal","Landsat","time series","California"],"languages":["en"],"rights":["Copyright 2016 Courtney Reents"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/90603","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Greenberg, Jonathan A."]},{"key":"dc:creator","label":"Author","values":["Reents, Courtney E"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2016-07-07T19:54:31Z","2016-04-21","2016-05"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Geography"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["forest disturbance","change detection","multitemporal","Landsat","time series","California"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2016 Courtney Reents"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/90603"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Natural and anthropogenic forest disturbances are a major influence on the global carbon cycle, with the local and global impacts of a disturbance event depending in large part on the timing, intensity, and cause of the disturbance. With some disturbance agents expected to increase in frequency and severity under the influence of climate change, knowledge of disturbance events and trends has become particularly crucial for a variety of scientific, political and management-related needs. Many studies have made use of time series analyses of Landsat imagery to study forest disturbance events at a variety of scales, but few have endeavored to attribute specific causal information to the disturbances detected, particularly at a subannual temporal scale. The purpose of this research is to investigate the suitability of a Landsat time series approach for detecting and describing the causes of disturbance events across the heterogeneous, forested landscapes of California. We used a random forest machine learning algorithm to relate normalized Landsat time series from 29,909 Landsat scenes extracted at the locations of known disturbance events with the cause of each event, including classes of logging, fire, pest damage, land use conversion, and no change. Validation of the model’s thematic accuracy performed using an independent dataset demonstrated an overall accuracy of 89.2% and a kappa value of 0.77, with by-class user’s accuracies ranging from 65% to 92% and by-class producer's accuracies ranging from 37% to 98%. We then applied this model to a sample footprint in California using all Landsat imagery available, generating a pixel-by-pixel disturbance record for every pixel from 1990 to 2010. While the model demonstrates some limitations in efficiency and temporal accuracy, it also highlights the capacity for disturbance detection at a subannual interval to offer insight into forest disturbance dynamics at a finer temporal scale.","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2016-07-07 without embargo terms","The student, Courtney Reents, accepted the attached license on 2016-04-21 at 09:05.","The student, Courtney Reents, submitted this Thesis for approval on 2016-04-21 at 09:25.","This Thesis was approved for publication on 2016-04-21 at 13:08.","DSpace SAF Submission Ingestion Package generated from Vireo submission #9394 on 2016-07-07 at 13:32:02","Made available in DSpace on 2016-07-07T19:54:31Z (GMT). No. of bitstreams: 3 REENTS-THESIS-2016.pdf: 3254776 bytes, checksum: d537d2c7c48b2e031d144c65fd92bfcd (MD5) courtney_reents_formatted_thesis_final.docx: 61185838 bytes, checksum: 6e4860c626ccb73d678aa512b319003f (MD5) LICENSE.txt: 4212 bytes, checksum: ccd936bc61fca4a9b1bb98a349a7160f (MD5) Previous issue date: 2016-04-21"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Detection and characterization of forest disturbances in California"]}]}],"canonical_facts":{"dc:contributor":["Greenberg, Jonathan A."],"dc:creator":["Reents, Courtney E"],"dc:date":["2016-07-07T19:54:31Z","2016-04-21","2016-05"],"dc:description":["Natural and anthropogenic forest disturbances are a major influence on the global carbon cycle, with the local and global impacts of a disturbance event depending in large part on the timing, intensity, and cause of the disturbance. With some disturbance agents expected to increase in frequency and severity under the influence of climate change, knowledge of disturbance events and trends has become particularly crucial for a variety of scientific, political and management-related needs. Many studies have made use of time series analyses of Landsat imagery to study forest disturbance events at a variety of scales, but few have endeavored to attribute specific causal information to the disturbances detected, particularly at a subannual temporal scale. The purpose of this research is to investigate the suitability of a Landsat time series approach for detecting and describing the causes of disturbance events across the heterogeneous, forested landscapes of California. We used a random forest machine learning algorithm to relate normalized Landsat time series from 29,909 Landsat scenes extracted at the locations of known disturbance events with the cause of each event, including classes of logging, fire, pest damage, land use conversion, and no change. Validation of the model’s thematic accuracy performed using an independent dataset demonstrated an overall accuracy of 89.2% and a kappa value of 0.77, with by-class user’s accuracies ranging from 65% to 92% and by-class producer's accuracies ranging from 37% to 98%. We then applied this model to a sample footprint in California using all Landsat imagery available, generating a pixel-by-pixel disturbance record for every pixel from 1990 to 2010. While the model demonstrates some limitations in efficiency and temporal accuracy, it also highlights the capacity for disturbance detection at a subannual interval to offer insight into forest disturbance dynamics at a finer temporal scale.","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2016-07-07 without embargo terms","The student, Courtney Reents, accepted the attached license on 2016-04-21 at 09:05.","The student, Courtney Reents, submitted this Thesis for approval on 2016-04-21 at 09:25.","This Thesis was approved for publication on 2016-04-21 at 13:08.","DSpace SAF Submission Ingestion Package generated from Vireo submission #9394 on 2016-07-07 at 13:32:02","Made available in DSpace on 2016-07-07T19:54:31Z (GMT). No. of bitstreams: 3 REENTS-THESIS-2016.pdf: 3254776 bytes, checksum: d537d2c7c48b2e031d144c65fd92bfcd (MD5) courtney_reents_formatted_thesis_final.docx: 61185838 bytes, checksum: 6e4860c626ccb73d678aa512b319003f (MD5) LICENSE.txt: 4212 bytes, checksum: ccd936bc61fca4a9b1bb98a349a7160f (MD5) Previous issue date: 2016-04-21"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/90603"],"dc:language":["en"],"dc:rights":["Copyright 2016 Courtney Reents"],"dc:subject":["forest disturbance","change detection","multitemporal","Landsat","time series","California"],"dc:title":["Detection and characterization of forest disturbances in California"],"dc:type":["text"],"thesis:degree_discipline":["Geography"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:26:34Z"}