.md .pdf

repository open issue suggest edit

Contents

8. Delayed Mode Quality Control

8.1. Dark count re calibration

As mentioned in Section 5, inter-calibrations between the gliders and CTD fluorescence sensors together with bottle samples of chlorophyll a are a fundamental tool in post recovery data processing. Additionally, to this, a further dark count correction is recommended when possible. Assuming no fluorescence signal in deeper waters, a deep-offset correction can be estimated. This is computed using the mean deep fluorescence (the deeper, the better) and can be subtracted from the data and used to further correct the results. This can be complex especially in study areas characterized by shallow waters, heavy resuspension, and overall complexity of the optical properties of the water column. Furthermore, when assessing this deep dark count offset, the contribution of deep sea red fluorescence on in situ dark counts should be determined. This can be done if the glider is equipped with reliable FDOM sensors (Xing et al., 2017).

8.2. Non photochemical quenching corrections

Fluorescence data has to be corrected for non-photochemical quenching (NPQ) in most cases. NPQ is a physiological response to high light environments used by plants and algae to protect themselves from damage and causes an evident weakening in fluorescence signal during the day. NPQ occurs only during the daytime or whenever light avaialbility is high, therefore when night measurements close in time and space are available, they can be used to correct daytime profiles. Different methods exist for NPQ the paper by (Thomalla et al., 2018) provides a good overview on the various methodologies. The most suitable methodology will vary based on the study area as well as the research question. The assumptions and corrections of each method may not provide credible results for the diffent water column structures so the methodology has to be chosen carefully.

Shall we add the nice table from the Thomalla paper summrizing all the methodologes? I add the draft below

Study	Equation	Assumptions
(Xing et al., 2012)	C1 = max0≤z≤MLD (Fl(z)) Flc(z) = C1; 0≤z≤d(C1)	- Fluorescence uniform within MLD - No quenching below max fluorescence within MLD
(Biermann et al., 2015)	C2 = max0≤z≤ED (Fl(z)) Flc(z) = C2; 0≤z≤d(C2)	- Fluorescence within ED uniform
(Swart et al., 2015)	C3 = max0≤z≤ED \(\left(\frac{Fl}{b_{bp}}\right)\) Flc(z) = C3 × bbp(z); 0≤z≤d(C3)	- fl : bbp ratio constant with depth
(Hemsley et al., 2015)	ChlNT = m × \(b_{{bp}_{NT}}\) + c* ChlDT(z) = m × bbp(z)DT + c; 0≤z≤ED *m (slope) and c (intercept)	- fl : bbp ratio constant with depth and time
(Thomalla et al., 2018)	FlcDT = \(\left(\frac{Fl_{NT}(z)}{Fl_{DT}(z)}\right)\) × bbp(z)DT; 0≤z≤QD If FlcDT(z) < FlcDT(z) then no correction is applied	- Same depth distribution of fl : bbp between night and day - No quenching at night

b_bp: particulate backscattering profile; d: depth; DT: daytime profile; ED: euphotic zone depth; Fl: fluorescence profile; Flc: corrected fluorescence profile; FlNT: averaged fluorescence over the night; MLD: mixed layer depth; NT: nighttime profile; QD: quenching depth; z: depth domain

We can say that if day and night profiles are used, the sunset/sunrise time can be used to separate the profiles or if PAR is available, this variable can used instead

previous

7. Post-recovery operations and calibrations

next

9. Data sharing