You are here:
Use of Satellite Remote Sensing to Improve Coastal Hypoxia Prediction
Citation:
Lehrter, J., D. Ko, C. Le, R. Gould, AND R. Dennis. Use of Satellite Remote Sensing to Improve Coastal Hypoxia Prediction. 2nd Gregory G. Leptoukh Online Giovanni Workshop, Online, November 10 - 12, 2014.
Impact/Purpose:
The purpose is to demonstrate how we have used satellite remote sensing products to improve our Gulf hypoxia models.
Description:
We describe the use of Giovanni satellite remote sensing products in the development and testing of a new modeling system that represents the processes leading to hypoxia (defined as water O2 concentration < 63 mmol m-3) on the Louisiana continental shelf (LCS). The modeling system consists of a biogeochemical model that is linked to a three-dimensional hydrodynamic model and implemented to study the effects of Mississippi River freshwater and nutrient loads on shelf carbon, nutrient, and oxygen cycles. The biogeochemical model is a new model called the Coastal General Ecosystem Model (CGEM) that is physically forced by our hydrodynamic model, the Navy Coastal Ocean Model as applied to the shelf (NCOM-LCS). We used the rainfall product from TRMM and ocean color products from MODIS Aqua/Terra and SeaWiFS to improve the surface salinity and heat fluxes to the model. Satellite sea surface temperature and altimetry products from multiple NASA and NOAA platforms were also directly assimilated into a Gulf-wide hydrodynamic model that supplied boundary conditions to NCOM-LCS. These improvements resulted in better representation of water-column vertical profiles of salinity and temperature, and, therefore, density stratification, which is a key physical process that sets the stage for hypoxia development. TRMM rainfall data were also used to improve estimates of atmospheric nitrogen loading to the sea surface by the Community Multi-Scale Air Quality (CMAQ) model. Ocean color data were used to improve CGEM’s representation of phytoplankton primary production, which is the process linking river nutrient loading to excess organic matter production to O2 depletion and hypoxia. Overall, the use of satellite data for boundary forcing, for data assimilation, and for calibration and verification of the modeling system resulted in improved model accuracy in comparison to observations. Future work will focus on data assimilation of ocean color products in order to synthesize these data in the model and to quantify the value these observations provide in terms of improving model predictions.