In a variety of remote sensing applications inverse modeling is used to retrieve parameters of ourenvironment from measurements of the radiance spectrum. As the dimensionality of the data increaseswith the number of spectral channels or as models become more complicated, inverse modeling becomes expensive in terms of computing time. We describe a technique to parameterize a given inverse model, which reduces the necessary computing time by orders of magnitude. An example that uses simulated data of SeaWiFS, an advanced ocean color sensor to be launched in 1993, is given, whereby the concentrations of suspended matter, phytoplankton, and gelbstoff (yellow substance) in water are determined.

© 1993 Optical Society of America

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Log in to access OSA Member Subscription

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Log in to access OSA Member Subscription

You do not have subscription access to this journal. Article level metrics are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Log in to access OSA Member Subscription