Guidelines for QC/QA from EMODNet Chemistry pilot

advertisement
Guidelines for QC/QA from EMODNet
Chemistry pilot
Marina Lipizer, Istituto Nazionale di Oceanografia e di Geofisica
Sperimentale–OGS Dipartimento OCE
Kick-off Meeting, June 3-4 2013, Trieste, Italy
EMODnet Pilot Project:
QA/QC: what has been done so far
Quality assurance is the end-to-end process, that ensures that the data
is captured, processed, analysed and harmonized according to a
predetermined set of conditions that are consistently applied
Quality control is the system of activities aimed to control the quality of
measurements so that it meets the needs of users. It can be applied at a
particular point during the process of acquiring (sampling-laboratory),
processing, analysing and harmonizing data
Why?
To ensure data consistency within a single data set and within a collection
of data sets and to ensure that the quality and errors of the data are
apparent to the user who has sufficient information to assess its suitability
for a task (IOC/UNESCO Manual, 1993).
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
EMODnet Pilot Project:
QA/QC: what has been done so far?
Challenge of 8 groups of compounds, 17 selected parameters in 3 matrices
“Survey” of existing QA/QC protocols
Identification of lacking QA/QC protocols
Definition of EMODnet QA/QC Guidelines
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
From: Chemical lot Pilot Project – Survey of existing QA/QC Protocols
Time 0
Matrices
Water
Parameter groups
✔ Fertilizers
✔
✔
✔
Sediment Biota
SeaDataNet
Partial pressure dissolved gases (e.g.O2)
Silicates
Chlorophyll
✔ Organic matter
✔
✔
✔
✔
✔
✔
✔
✔
Acidity
Heavy metals
Hydrocarbons
Pesticides & biocides
Antifoulants
Pharmaceuticals
Radionuclides
Identification of lacking QA/QC protocols
Plastics
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Water
For: Temperature, Salinity, fertilizers,
dissolved oxygen
From SDN QC Protocols:
AUTOMATIC QUALITY CONTROL CHECKS:
On Metadata:
Date and time (valid format)
Latitude and longitude (valid format)
Position must not be on land!
On Data:
Global range (expected extremes encountered in the oceans)
Regional range (expected extremes encountered in particular regions)
Deepest pressure (…profile does not contain pressures higher than the bottom
bathymetry)
Spike (…large differences between adjacent values)
Gradient (vertically adjacent salinity and temperature data not too different)
Density inversion (…density normally increasing with increasing pressure in a
profile)
Pressure increasing (Pressures from the profile monotonically increasing).
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Water
For: Temperature, Salinity, fertilizers,
dissolved oxygen
From SDN QC Protocols:
“SCIENTIFIC/VISUAL” QUALITY CONTROL CHECKS:
On Data (with important contribution from Regional Experts):
Property-property plot (e.g. TS, N:P, nutrients-salinity,...)
Compare with climatology
Comparison with min-max ranges for the regions
Tools used: ODV visual inspection (broad range check)
Example from the Adriatic Sea:
Nutrients in the water column:
with the use of ODV tool: outliers detection.
Surface distribution
Vertical distribution
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Definition of EMODnet QA/QC Guidelines
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Biota
➢ Species (ERMS or ITIS)
➢ Number of individuals
➢ Tissue
Data guidelines
General to all matrices
➢
➢
➢
➢
➢
➢
➢
➢
➢
Missing values
Use of CAS numbers
Intercalibrations
Reference Materials
Uncertainty
Code list
Cruise ID's
Basis (of determination:i.e.specific matrix)
Depth
Meta-data requirements
➢
➢
➢
➢
➢
➢
➢
Purpose of monitoring
Station names
Sampling method and instrument
Method of analysis
Method of filtration for trace metals
Method of chemical extraction
Limit of detection/limit of determination
Water
➢ Method of filtration for trace metals
➢ Suspended Particulate Matter
Sediment
➢ Method of chemical extraction
➢ Grain size
➢ Normalisers
Controlled Vocabularies
➢ Parameters
➢ Units
Normalisation (by sampling
matrix)
➢ Sediment
➢ Biota
➢ Water
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Definition of EMODnet QA/QC Guidelines
Further outcomes:
Guidelines for:
Fertilizers and Dissolved
gas (Water) – reference for
Black Sea
Guidelines for:
Antifoulants (Biota)
Guidelines for:
Hydrocarbons, pesticides, heavy
metals (Water & Sediment)
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
http://www.emodnet-chemistry.eu/portal/portal/emodnet/Extranet
From: Chemical
lot Pilot Project
user name: emduser
password: 91L07G207
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
From: Chemical lot Pilot Project – Survey of existing QA/QC Protocols
Time =End of Pilot Project
Matrices
Water
Parameter groups
✔ Fertilizers
SeaDataNet
✔ Silicates
✔ Partial pressure dissolved gases (e.g.O2)
✔ Chlorophyll
Sediment Biota
EMODnet Guidelines
✔ Organic matter
✔ Acidity
✔
✔
✔
✔
✔
✔
Heavy metals
Hydrocarbons
Pesticides & biocides
Antifoulants
Pharmaceuticals
Radionuclides
EMODnet Regional QA/QC Procedure (BS)
✔ Plastics
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
ICES-Report 47
Open issues form Chemical lot PP
For “Exotic (not SDN)” parameters and contaminants it was
highlighted that:
➔Difficult to fit SDN classic QC protocol!
➔Sediment and biota (missing in SDN)
➔How to apply Range checks? Not enough data availability for a climatology;
➔How to apply Spike checks? The contaminants are more related to an “event” logic…
high-values could be related to events, not necessary to errors (sampling/analytical)!
➔Often data below detection limits… (not specified!!!)
…some of the lessons learned…
The complexity of the measurements covering 8 groups of parameters collected on 3
matrices → need to have improved metadata description and continue with adapting
process of SDN standards;
The “exotic parameters” → need an ad hoc QC protocol (no spikes, difficult to apply
ranges);
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Future:
Verify MSFD requirements
Mantain and possibly implement SDN Protocols (for other matrices,
contaminants)
Improve inventory of existing QA/QC Protocols (any others?)
Hopefully new data harvesting will supply enough data to calculate ranges
for exotic parameters;
Involvement of all Marine Conventions highly required, in particular
increase contact with Mediterranean and Black Sea (MEDPOL, Black Sea
Convention...) to obtain local ranges;
Consider proposal of a 2 levels QF scheme from IODE workshop
(http://www.iode.org/index.php?option=com_oe&task=viewDocumentRecord
&docID=5109)
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
EMODnet 2:
WP3: QA/QC – Validation – MSFD Interaction
WP Leader: ISPRA (Italy)
➢ Guidelines with MSFD requirements
➢ Quality Control and Quality Assurance
➢ Inventory of analytical methods
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
New big challenge:
Inventory of analytical methods
Why?
To collect detailed information on analytical method, uncertainty and
detection/quantification limit for each parameter and different datasets
Particular requirements:
Limit of detection for synthetic pollutants (pesticides, antifoulants, PAHs,...) (is zero
really absence of contamination?) (Background concentration of synthetic
pollutants has to be zero, but sometimes it is “below detection limit” (is this stated?)
To better assess intercomparison among datsets (tests with certified reference
standards? Long term data consistency?)
Expected outcomes:
Inventory of analytical method descriptions for sample analysis
…..High quality data products published!
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Data of known quality
are more useful
than data of unknown quality
Thanks for your attention!
M. Lipizer - Guidelines for QC/QA from EMODNet Chemistry pilot
Download