Abstract's details
Sea Level ECV quality assessment via global Ocean model assimilation
CoAuthors
Event: 2014 Ocean Surface Topography Science Team Meeting
Session: Quantifying Errors and Uncertainties in Altimetry data
Presentation type: Type Oral
Contribution: PDF file
Abstract:
In the ocean modeling community satellite data, especially SSH fields, are assimilated on a regular basis. SSH fields are very important in this context because of their dynamical relevance for constraining the ocean's flow field. However, assimilating SSH data into an ocean model does not only improve the quality of model but in addition, can also help testing the quality and the consistency of the input data as well.
In our work we aim to quantify improvements in Sea Level (SL) data through the ESA - Climate Change Initiative (CCI) effort and we aim to test the consistency of the Essential Climate Variable (ECV) of Sea Level (SL_ECV) with other ECVs through the assimilation process and to investigate where remaining inconsistencies exist and why.
For this purpose the GECCO2 assimilation approach assimilates SL_ECV jointly with all other available ECVs over the ocean and in situ data. The dynamically consistent ocean state estimation adjusts only uncertain model parameters to bring the model into consistency with ocean observations. Improvements in data products can be investigated by studying the residuals between the different data products and the constrained model. With this approach we can demonstrate that in many regions the SL_ECV has been improved from version V0 to version V1. However, there are regions where SL_ECV_V1 is further away from the model "truth". In that sense it is important to understand that the model assimilated SL_ECV_V0 and therefore has tried to adapt to the SL_ECV_V0. Therefore, inconsistencies existed when comparing the synthesis results to the updated version SL_ECV_V1! These deviations between the model "truth" and the improved data product (SL_ECV_V1) increased mostly in low energetic areas. To quantify these deviations, we perform further assimilation runs with the updated data product SL_ECV_V1.1 as well. Hence we will be able to explain the deviations from the model "truth" for the version V0 assimilation run. Especially in the low energetic regions we expect improved residuals after the new assimilation runs that are using SL_ECV_V1.1.
In our work we aim to quantify improvements in Sea Level (SL) data through the ESA - Climate Change Initiative (CCI) effort and we aim to test the consistency of the Essential Climate Variable (ECV) of Sea Level (SL_ECV) with other ECVs through the assimilation process and to investigate where remaining inconsistencies exist and why.
For this purpose the GECCO2 assimilation approach assimilates SL_ECV jointly with all other available ECVs over the ocean and in situ data. The dynamically consistent ocean state estimation adjusts only uncertain model parameters to bring the model into consistency with ocean observations. Improvements in data products can be investigated by studying the residuals between the different data products and the constrained model. With this approach we can demonstrate that in many regions the SL_ECV has been improved from version V0 to version V1. However, there are regions where SL_ECV_V1 is further away from the model "truth". In that sense it is important to understand that the model assimilated SL_ECV_V0 and therefore has tried to adapt to the SL_ECV_V0. Therefore, inconsistencies existed when comparing the synthesis results to the updated version SL_ECV_V1! These deviations between the model "truth" and the improved data product (SL_ECV_V1) increased mostly in low energetic areas. To quantify these deviations, we perform further assimilation runs with the updated data product SL_ECV_V1.1 as well. Hence we will be able to explain the deviations from the model "truth" for the version V0 assimilation run. Especially in the low energetic regions we expect improved residuals after the new assimilation runs that are using SL_ECV_V1.1.