Show simple item record

dc.contributor.advisorEriksson, Marian
dc.creatorAustin, Michael Andrew
dc.date.accessioned2017-03-02T16:48:01Z
dc.date.available2017-03-02T16:48:01Z
dc.date.created2016-12
dc.date.issued2016-11-28
dc.date.submittedDecember 2016
dc.identifier.urihttps://hdl.handle.net/1969.1/159025
dc.description.abstractThe Public Land Survey System (PLSS) serves as a means of legally identifying land in 30 out of 50 U.S. states. PLSS coordinates are frequently used to describe spatial data such as the locations of wildfires, but PLSS coordinates are largely incompatible with most software for spatial analysis. One means of translating PLSS coordinates into longitude and latitude had been a regression-based tool known as TRS2LL which was developed by Martin Wefald, but the model on which this tool is based only supported 17 out of 30 PLSS states and the geographic boundaries of the regions for which each regression was applied to (regression domains) had been delineated by hand on paper maps. We present a new model known as PLSS2LL based loosely on TRS2LL in which regression domain boundaries are procedurally generated via GIS and the coverage is extended to all 30 PLSS states. We observed an improvement in PLSS2LL’s accuracy (mean error 133.62 m) in predicting longitude and latitude coordinates over its predecessor TRS2LL (mean error 220.58 m). While more accurate, the resulting domains are more fragmented with an average of about 280 domains for each of 30 states, 8,439 total as compared to Wefald’s average of 105 for each of 17 states. Approximately 59.60% of PLSS2LL’s predictions fell within 100 m, versus the 29.41% of TRS2LL’s. The inverse conversions in which longitude and latitude are used to predict PLSS coordinates were also tested. For predicting PLSS coordinates, PLSS2LL and TRS2LL respectively yielded accuracies of 90.33% and 84.48%. Our model effectively reduced the original data from the U.S. Bureau of Land management by factor of 143.75. We also explored the various sources of error and considered future improvements.en
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.subjectPLSSen
dc.subjectlongitudeen
dc.subjectlatitudeen
dc.subjectconversionen
dc.titleConversion of PLSS Coordinates Using Automatically Generated Regression Domainsen
dc.typeThesisen
thesis.degree.departmentEcosystem Science and Managementen
thesis.degree.disciplineEcosystem Science and Managementen
thesis.degree.grantorTexas A & M Universityen
thesis.degree.nameMaster of Scienceen
thesis.degree.levelMastersen
dc.contributor.committeeMemberKlein, Andrew
dc.contributor.committeeMemberPopescu, Sorin
dc.type.materialtexten
dc.date.updated2017-03-02T16:48:01Z
local.etdauthor.orcid0000-0001-8597-395X


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record