Nitrate - Easy Q.C.V BGC

The Nitrate calibration methodology follows the Argo reference documentation BGC-Argo quality control manual for nitrate concentration and recommendation from Euro-Argo RISE project D4.5: Recommendations for enhancement of NO3 QC Methods.

Briefly, this consists of 4 main steps:

Step 1: Selection of the equation for the adjustment computation
Step 2: Selection of the method to assess the reference value of Nitrate
Step 3: Selection of the methodologies for evaluating referenced pressure P
Step 4: Definition of the number of linear regression / segment in time

Step 1: Selection of the equation for the adjustment computation¶

Two sets of equation(s) are possible:

Method 1: OFFSET and DRIFT

The method calculates at depth (at the reference pressure Pref, where there should be the minimum of the variability) the difference between the measurements and reference data and fits this difference as a linear function of time. When needed, the linear function can be estimated by segments of time. The same correction estimated at Pref is used along the whole vertical profile.

nitrate\_adjusted_{(c,P)} = nitrate_{raw(c,P)} - correction_{(c,Pref)}

(1)

correction_{(c,Pref)} = offset_{(i,Pref)} + drift_{(i,Pref)}*(time_c -time_i)/365

(2)

Where :

offset (in µmol/kg) and drift (in µmol/kg/year) are respectively the intercept and the slope of the linear regression of $nitrate_{raw_(P=Pref)} - nitrate_{ref(P=Pref)}$ as a function of time expressed in julian day.
c corresponds to the cycle vertical profile (in Argo, the vertical profile is performed once every 10 days for the usual mission configuration)
i corresponds to the time segment number if necessary - i.e. if the slope value changes by part during the lifetime of the sensor.
Pref is set equal correspond to the referenced pressure
time_i corresponds to the time (in julian day) when the segment i starts.

Step 2: Selection of the method to assess the reference value of Nitrate¶

Several methodologies are available to assess nitrate_ref. They all have their specificities, pros and cons. The B.G.C QCV proposed the methodologies recommended by the Euro-Argo RISE D4.5: Recommendations for enhancement of NO3 QC Methods⁠ :

Neural network prediction system
- CANYON-B
- CANYON-MED (Mediterranean Sea)
Gridded climatologies
- World Ocean Atlas monthly climatology
- World Ocean Atlas decadal climatology

Step 3: Selection of the methodologies for evaluating referenced pressure P¶

Two options are possible:

From nitrate measured by the platform (minimum variability zone)
From the operator

Diagnostic figures are made by the tool for explaining its choice.

Diagnostic figure examples!

From nitrate measured by platform

From operator

Figure 1:Z reference for the nitrate calibration estimated by the tool (red)

Step 4: Definition of the number of linear regression / segment in time¶

This option defines the number offset + drift that should be used for estimating the adjustment. Three options are possible :

Automatically defined by the tool
defined by the user
favors one linear regression

Diagnostic figures are made by the tool for explaining its choice.

Diagnostic figure examples!

Figure 3:Single and Multiple Linear Regression defined by the tool

References¶

Johnson, K., Maurer, T., Plant, J., Bittig, H., Schallenberg, C., & Schmechtig, C. (2021). BGC-Argo quality control manual for nitrate concentration. Ifremer. 10.13155/84370

Easy Q.C.V. BGC workflows

General description

Easy Q.C.V. BGC workflows

Chlorophyll-a