Quality control of ε estimates (QA2): Difference between revisions

From Atomix
← Older editNewer edit →
VisualWikitext
Yuengdjern (talk | contribs)
215 edits
No edit summary
No edit summary
Line 1: Line 1:
# Reject ensembles where the slope is negative (i.e. ε<sub>i</sub><0)
# Data segments for which the regression coefficient a<sub>1</sub> '''[LINK TO PREVIOUS PAGE]''' is negative yield an imaginary <math>\varepsilon</math> value, which should be rejected
# Reject ensemble where Ni values are unreasonable (I.e. N<sub>Subscript text</sub>i<0).
# Data segments for which the regression coefficient a<sub>0</sub> '''[LINK TO PREVIOUS PAGE]''' is negative (implying a negative noise floor) are likely to be invalid and are typically rejected
# Estimate the uncertainty in ε<sub>i</sub> as the 95% confidence interval on the slope. Reject values that are too large (e.g. delta ε<sub>i</sub> / ε<sub>i</sub> >0.6).  
# Examine the consistency of <math>\varepsilon</math> between bins (if evaluated) and between beams as an indication of estimate reliability - the geometric mean between beams is frequently used as the representative value
# Average the ε<sub>i</sub> estimates from the diverging beams (if applicable) to determine ε.
# Evaluate the impact of varying r<sub>max</sub> values (within the anticipated inertial range) on <math>\varepsilon</math>; an increase in <math>\varepsilon</math> with increasing r<sub>max</sub> is likely to indicate that v’ retains a non-turbulent contribution to the velocity difference between bins
# Check if distribution of ε is lognormal
# The goodness of fit (R<sup>2</sup>) for the regression provides a basic indication of the quality of the fit
# A better indication of the quality of the fit is usually  provided by looking at the ratio of the estimated <math>\varepsilon</math> value to that based on the 95%-ile confidence interval estimate of the a<sub>1</sub> regression coefficient e.g. reject values where the ratio exceeds a specified threshold
# Examine the distribution of <math>\varepsilon</math> estimates - in most situations, this would be expected to be log-normal
# Comparison of observed values with nominal values based on established boundary-forced scalings may also be informative and help to identify observation or processing issues


   
   
Return to [[ADCP structure function flow chart| ADCP Flow Chart front page]]
Return to [[ADCP structure function flow chart| ADCP Flow Chart front page]]

Revision as of 14:37, 10 November 2021

  1. Data segments for which the regression coefficient a1 [LINK TO PREVIOUS PAGE] is negative yield an imaginary [math]\displaystyle{ \varepsilon }[/math] value, which should be rejected
  2. Data segments for which the regression coefficient a0 [LINK TO PREVIOUS PAGE] is negative (implying a negative noise floor) are likely to be invalid and are typically rejected
  3. Examine the consistency of [math]\displaystyle{ \varepsilon }[/math] between bins (if evaluated) and between beams as an indication of estimate reliability - the geometric mean between beams is frequently used as the representative value
  4. Evaluate the impact of varying rmax values (within the anticipated inertial range) on [math]\displaystyle{ \varepsilon }[/math]; an increase in [math]\displaystyle{ \varepsilon }[/math] with increasing rmax is likely to indicate that v’ retains a non-turbulent contribution to the velocity difference between bins
  5. The goodness of fit (R2) for the regression provides a basic indication of the quality of the fit
  6. A better indication of the quality of the fit is usually provided by looking at the ratio of the estimated [math]\displaystyle{ \varepsilon }[/math] value to that based on the 95%-ile confidence interval estimate of the a1 regression coefficient e.g. reject values where the ratio exceeds a specified threshold
  7. Examine the distribution of [math]\displaystyle{ \varepsilon }[/math] estimates - in most situations, this would be expected to be log-normal
  8. Comparison of observed values with nominal values based on established boundary-forced scalings may also be informative and help to identify observation or processing issues


Return to ADCP Flow Chart front page

Retrieved from "http://atomix.app.uib.no/index.php?title=Quality_control_of_ε_estimates_(QA2)&oldid=2079"