Incorporating Uncertainties in Satellite-Derived Chlorophyll into Model Forecasts
Abstract
We describe and apply an ensemble approach, similar to that used in environmental modeling, to quantify errors and produce uncertainty maps for satellite-derived ocean color chlorophyll, and we incorporate these uncertainties into hydrodynamic and biophysical models. For an ocean color image, we first apply realistic noise to the satellite top-of-atmosphere radiances, which leads to an ensemble of chlorophyll images. From this ensemble, we derive mean and standard deviation (uncertainty) images for the chlorophyll, which we then incorporate into both hydrodynamic and biophysical forecast models. For both these cases, we create forecast ensemble suites; the ensemble variance provides an indication of uncertainty, or confidence in the chlorophyll forecast. We examine mean and individual forecast ensemble members (R2, spread-skill statistics) to assess predictive value. Thus, we produce a final chlorophyll forecast field that includes uncertainties in both the initial satellite chlorophyll values as well as uncertainties in the hydrodynamic and biological models.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2012
- Accession Number
- ADA571870
Entities
People
- E. Coelho
- I. Shulman
- P. Sakalaukas
- R. W. Gould Jr.
- S. C. Mccarthy
- Steven P. Anderson
Organizations
- United States Naval Research Laboratory