QC%20Procedures%20at%20INCOIS

1
QC Procedures at INCOIS

• TVS Udaya Bhaskar,
• CLIVAR(GSOP), 12 - 14th June, 2013

2
Type of Data sets

• Argo data from Indian and non-Indian floats.
• CTD data from Indian ORV.
• XBT/XCTD data from National Institute of
  Oceanography (NIO), Goa.
• Specific Project related data viz., ARMEX,
  BOBMEX, ICRP etc.
• Time series data also available.
• Moorings, Drifters, Current Meter Arrays, ADCPs
  etc.

3
QC procedures

• For CTDs and Argo floats
• Automatic QC as proposed by ADMT
• Three-way-QC
• Automatic QC,
• QC based on Objective Analysis
• Visual Quality Control Tool
• Cluster based
• Latitude, Longitude based clusters
• For XBT/XCTDs
• QC is performed at NIO, Goa
• Planning to use Mquest by CSIRO.

4
(No Transcript)
5
Visual Quality Control of Argo/CTD data

• Visual Quality Control Tools is used for QC of
  Argo data. WOA05 is used in the background for
  the QC process.
• Profiles falling out of 3sigma are set as bad.

TS Profiles (Argo Floats, XBT, XCTD, CTD)
6
Quality control of oceanographic data sets based
on spatio-temporal clustering

• Clustering method with recent WOA09 climatology.
• Tested for all standard levels
• Sensitivity studies were done.

7
QC based on Lat/Lon Clusters

• Climatological mean and standard deviation are
  used to build n-sided (dimensional) polygon.
• Outliers are detected based by judging whether
  the points are falling with in the polygon or out
  side.
• These are re-examined for any special events and
  then injected into the database.

8
Good point
Outlier
The n-sided polygon is used for checking
outliers. Now the data points whose quality is to
checked are introduced and checked whether the
points fall with in the n-sided polygon (good
points) or outside the polygon (bad points).
9
Depth 10 m
Outliers
Outliers
Latitude (deg)
10
Persian Gulf
Red Sea
Outliers
BOB
11
(No Transcript)
12
Thank You