DUACS DT2014: the new multi-mission altimeter data set reprocessed over 20 years
The new DUACS DT2014 reprocessed products have been available since April 2014. Numerous innovative changes have been introduced at each step of an extensively revised data processing protocol. The use of a new 20-year altimeter reference period in place of the previous 7-year reference significantl...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Copernicus Publications
2016-09-01
|
Series: | Ocean Science |
Online Access: | http://www.ocean-sci.net/12/1067/2016/os-12-1067-2016.pdf |
Summary: | The new DUACS DT2014 reprocessed products have been available since April
2014. Numerous innovative changes have been introduced at each step of an
extensively revised data processing protocol. The use of a new 20-year
altimeter reference period in place of the previous 7-year reference
significantly changes the sea level anomaly (SLA) patterns and thus has a
strong user impact. The use of up-to-date altimeter standards and geophysical
corrections, reduced smoothing of the along-track data, and refined mapping
parameters, including spatial and temporal correlation-scale refinement and
measurement errors, all contribute to an improved high-quality DT2014 SLA
data set. Although all of the DUACS products have been upgraded, this paper
focuses on the enhancements to the gridded SLA products over the global
ocean. As part of this exercise, 21 years of data have been homogenized,
allowing us to retrieve accurate large-scale climate signals such as global
and regional MSL trends, interannual signals, and better refined mesoscale
features.<br><br>An extensive assessment exercise has been carried out on this data set, which
allows us to establish a consolidated error budget. The errors at mesoscale
are about 1.4 cm<sup>2</sup> in low-variability areas, increase to an average of
8.9 cm<sup>2</sup> in coastal regions, and reach nearly 32.5 cm<sup>2</sup> in high
mesoscale activity areas. The DT2014 products, compared to the previous
DT2010 version, retain signals for wavelengths lower than ∼ 250 km,
inducing SLA variance and mean EKE increases of, respectively, +5.1 and
+15 %. Comparisons with independent measurements highlight the improved
mesoscale representation within this new data set. The error reduction at the
mesoscale reaches nearly 10 % of the error observed with DT2010. DT2014
also presents an improved coastal signal with a nearly 2 to 4 % mean
error reduction. High-latitude areas are also more accurately represented in
DT2014, with an improved consistency between spatial coverage and sea ice
edge position. An error budget is used to highlight the limitations of the
new gridded products, with notable errors in areas with strong internal
tides. |
---|---|
ISSN: | 1812-0784 1812-0792 |