Show simple item record

dc.contributor.advisorRivera-Gallego, Wilson
dc.contributor.authorGerardino-Neira, Carolina
dc.date.accessioned2019-05-15T17:59:33Z
dc.date.available2019-05-15T17:59:33Z
dc.date.issued2007
dc.identifier.urihttps://hdl.handle.net/20.500.11801/2363
dc.description.abstractThis research presents a sensitivity analysis of a semi-analytical inversion model for hyperspectral remote sensing of shallow coral ecosystems. Using this inversion model, five parameters describing water column bioptical properties, bathymetry and magnitude of bottom reflectance are retrieved. In addition to the parameters of interest, the model contains 12 nuisance parameters that are traditionally assigned a fixed set of values. A sensitivity analysis of estimates retrieved to these nuisance parameters is accomplished using SimLab software to study their impact on model output. The computationally intensive analysis was enabled implementing the inver- sion model within a parallel processing framework using GENCAN. The sensitivity analysis was used to identify which nuisance parameters are most influential on the parameters of interest. The nuisance parameters found to be most relevant are: S, the spectral slope of the absorption coefficient for gelbstoff, Y, the spectral power coefficient for calculating the backscattering coefficient, and Dop, a constant in the equation for the distribution function for scattered photons from the bottom.en_US
dc.description.abstractEsta investigación presenta un análisis de sensibilidad aplicado a un modelo semi-analítico inverso, para imágenes hiperespectrales de ecosistemas de aguas poco profundas. El modelo de inversión encuentra cinco parámetros de interés los cuales describen propiedades bio-ópticas de la columna de agua, profundidad y reflectancia del fondo marino. Adicionalmente, el modelo tiene 12 parámetros nuisance, cuyos valores son tradicionalmente fijados. Se utilizó el software SimLab para un análisis de sensibilidad de los estimados de los parámetros nuisance y ver su impacto en la salida del modelo. El complejo análisis computacional fue posible implementando el modelo de inversión en una infraestructura para procesamiento en paralelo utilizando el método GENCAN. El análisis de sensibilidad permitió encontrar los parámetros nuisance más influyentes en los parámetros de interés. Los parámetros nuisance más importantes son: S, la pendiente espectral del coeficiente de absorción para gelbstoff, Y, la potencia espectral del coeficiente de retrodispersión, y Dop, constante en la función de distribución por fotones dispersos desde el fondo.en_US
dc.description.sponsorshipSupported in part by CenSSIS, the Center for Subsurface Sens- ing and Imaging Systems, under the Engineering Research Centers Program of the National Science Foundation (Award Number ECE-986821). This work was also supported by NASA Training Grant NNG05GG78H (PR Space Grant) and NASA Cooperative Agreement NNX07AO30A (PR NASA EPSCoR).en_US
dc.language.isoEnglishen_US
dc.titleUtilizing high-performance computing to improve performance and investigate sensitivity of an inversion model for hyperspectral remote sensing of shallow coral ecosystemsen_US
dc.typeThesisen_US
dc.rights.licenseAll rights reserveden_US
dc.rights.holder(c) 2007 Carolina Gerardino-Neiraen_US
dc.contributor.committeeVelez-Reyes, Miguel
dc.contributor.committeeGoodman, James
dc.contributor.representativeColón-Reyes, Omar
thesis.degree.levelM.S.en_US
thesis.degree.disciplineElectrical Engineeringen_US
dc.contributor.collegeCollege of Engineeringen_US
dc.contributor.departmentDepartment of Electrical and Computer Engineeringen_US
dc.description.graduationYear2007en_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

  • Theses & Dissertations
    Items included under this collection are theses, dissertations, and project reports submitted as a requirement for completing a graduate degree at UPR-Mayagüez.

Show simple item record

All rights reserved
Except where otherwise noted, this item's license is described as All Rights Reserved