optimalSensor.evaluate

Evaluates the variance reduced by individual proxies in a network


Syntax



Description

deltaVariance = obj.evaluate

For each observation site, evaluates the amount of the metric’s variance that is reduced when the site is assimilated alone. Evaluates the individual influence of each proxy observation on an assmilation.

Important This method neglects the covariance between observation sites, and only considers the effects of assimilating each site individually. In general, when observation estimates covary, their total effect on the metric’s variance will be smaller than the sum of their individual effects. Use the “optimalSensor.update” method to assess the total effect on metric variance when multiple estimates covary. Also note that we are not referring to R uncertainty covariances here. This effect still occurs when using R uncertainty variances.


Output Arguments

deltaVariance

numeric vector [nSites]
The reduction of variance that occurs when assimilating each site individually.