Accounting for Input Noise in Gaussian Process Parameter Retrieval
This addresses error propagation in remote sensing for geoscientists, but it is incremental as it extends an existing GP framework.
The paper tackled the problem of Gaussian processes (GPs) assuming no input noise, which is unrealistic in earth observation, by developing a method to account for input noise using error propagation through the derivative of the predictive mean. The result showed more accurate model error representation in a temperature prediction problem from infrared sounding data.
Gaussian processes (GPs) are a class of Kernel methods that have shown to be very useful in geoscience and remote sensing applications for parameter retrieval, model inversion, and emulation. They are widely used because they are simple, flexible, and provide accurate estimates. GPs are based on a Bayesian statistical framework which provides a posterior probability function for each estimation. Therefore, besides the usual prediction (given in this case by the mean function), GPs come equipped with the possibility to obtain a predictive variance (i.e., error bars, confidence intervals) for each prediction. Unfortunately, the GP formulation usually assumes that there is no noise in the inputs, only in the observations. However, this is often not the case in earth observation problems where an accurate assessment of the measuring instrument error is typically available, and where there is huge interest in characterizing the error propagation through the processing pipeline. In this letter, we demonstrate how one can account for input noise estimates using a GP model formulation which propagates the error terms using the derivative of the predictive mean function. We analyze the resulting predictive variance term and show how they more accurately represent the model error in a temperature prediction problem from infrared sounding data.