A new automatic quality control system for ocean profile observations
and impact on ocean warming estimate
The rapidly enhancing global hydrographic in-situ observations archive
is data quality heterogeneous. Different data applications (e.g.,
climate change science) require a high-performance quality control (QC)
system to reliably identify outliers in profile data. This study
presents a new automatic QC procedure (CAS-Ocean Data Center quality
control system; CODC-QC) for ocean in-situ temperature outliers
detection. Unlike many existing QC procedures, no assumption is made of
a Gaussian distribution law in CODC-QC as the oceanic variables are
typically skewed. Instead, the 0.5% and 99.5% quantiles are used to
define the local temperature climatological ranges. Additionally, we
constructed local climatological ranges for the vertical temperature
gradient which increased the ability of identifying spurious profiles.
The performance of CODC-QC was evaluated using two benchmark datasets.
Results demonstrated that CODC-QC is effectively in removing spurious
data and minimizing the percentage of mistakenly flagged good data.
Additionally, the CODC-QC was applied to the global World Ocean Database
(WOD) historical temperature profiles and a significant
quality-dependent on instrumentation types was found. Finally, as ocean
heat content (OHC) is a fundamental indicator of climate change, the
impact of different QC systems on OHC estimation is examined. Results
based on an existing mapping approach indicate that applying CODC-QC
system leads to a 41.7 % (4.9%) difference for linear trend of the
global 0-2000m OHC changes within 1955-1990 (1991-2020) compared to the
WOD-QC, implying a non-negligible source of error in ocean warming
estimate. The new QC procedure could support further improvement of the
oceanic climate records and other applications.