Difference between revisions of "Data Quality and Validation White Paper Kick-off, GSFC"

From Earth Science Information Partners (ESIP)
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
 +
 +
Participants: Joanne Nightingale, Chris Lynnes, Greg Leptoukh
 +
 
==White Paper on Data Quality and Validation Framework for RS data: Best practices and user perspective==
 
==White Paper on Data Quality and Validation Framework for RS data: Best practices and user perspective==
1. "Objective":
+
1. ''Objective'':
 
Capture, harmonize and provide useful data quality for RS data
 
Capture, harmonize and provide useful data quality for RS data
  
2. "Current Status":
+
2. ''Current Status'':
 
* No coordinate approach ... besides QA4EO, WGCV and some discipline-specific efforts
 
* No coordinate approach ... besides QA4EO, WGCV and some discipline-specific efforts
 
* Duplication of efforts across missions - inefficient utilization of funding
 
* Duplication of efforts across missions - inefficient utilization of funding
Line 11: Line 14:
 
* Fitness for purpose: quality need should depend on data usage but the current quality is based on validation that has the specific purpose of validating instrument and retrieval algorithms
 
* Fitness for purpose: quality need should depend on data usage but the current quality is based on validation that has the specific purpose of validating instrument and retrieval algorithms
  
3. "Suggested approach":
+
3. ''Suggested approach'':
 
* Collect best practices from various communities and known efforts
 
* Collect best practices from various communities and known efforts
 
* Develop consistent terminology and metrics, e.g., completeness, consistency, representativeness
 
* Develop consistent terminology and metrics, e.g., completeness, consistency, representativeness
* Identify main purpose classes and establish a high-level Q-metrics levels for different purposes
+
* Identify main purpose classes and establish a high-level Q-metrics levels for different purposes:
*
+
** Climate change; model validation; event monitoring; tele
 +
* Error budget:
 +
** Propagate cal/L1 errors to L2 and L3
 +
* Bias assessment:
 +
** Quantify Spatial/temporal/vertical sampling effect
 +
* Provenance:
 +
** Capture, harmonize and deliver useful provenance
 +
 
 +
4. ''Decadal Survey era''

Latest revision as of 16:01, December 22, 2010

Participants: Joanne Nightingale, Chris Lynnes, Greg Leptoukh

White Paper on Data Quality and Validation Framework for RS data: Best practices and user perspective

1. Objective: Capture, harmonize and provide useful data quality for RS data

2. Current Status:

  • No coordinate approach ... besides QA4EO, WGCV and some discipline-specific efforts
  • Duplication of efforts across missions - inefficient utilization of funding
  • Validation:
    • L2: validate in some areas and extrapolate globally. Issues: filtering by QC flag doesn't necessarily lead to good product
    • L3: what is L3 validation?
  • Fitness for purpose: quality need should depend on data usage but the current quality is based on validation that has the specific purpose of validating instrument and retrieval algorithms

3. Suggested approach:

  • Collect best practices from various communities and known efforts
  • Develop consistent terminology and metrics, e.g., completeness, consistency, representativeness
  • Identify main purpose classes and establish a high-level Q-metrics levels for different purposes:
    • Climate change; model validation; event monitoring; tele
  • Error budget:
    • Propagate cal/L1 errors to L2 and L3
  • Bias assessment:
    • Quantify Spatial/temporal/vertical sampling effect
  • Provenance:
    • Capture, harmonize and deliver useful provenance

4. Decadal Survey era