DMQC for Bio-Argo Data

Towards Delayed-Mode Quality Control for Bio-Argo Data

Corresponding author : Hervé Claustre 

  

Context

A new array of biogeochemical measurements obtained from Argo autonomous profiling floats arises. Following the pathway indicated by the physical oceanographers, networks of biogeochemical profiling floats (BPF) are under construction (see figure 1), with the specific and declared aim to provide the involved scientific community with a continuous, real-time, and automatic flux of accurate biogeochemical observations (Claustre et al., 2010). BPF networks will assure, in the future, the necessary data flow to initialise, constrain and validate operational real-time ecosystem models. Such an automated «bio» platforms array will also provide a very useful biological information to allow the future extraction of climatic trends.

A new series of recent technological improvements have led to the fabrication of miniaturized and very low energy demanding biogeochemical sensors. But conversely to what happens when the same kinds of equipments are operated from a ship, these bio-optical data are collected from Argo floats in environmental conditions which are out of the operator’s control. Calibration factors and characterizations provided by the manufacturer are not sufficient to ensure the stability of the biogeochemical sensors during the 2-3 years lifetime of the autonomous platforms. For the next years, the challenge is thus to achieve to deliver, in an automatic way, quality-controlled and consistant biogeochemical data.

The aim of the present study is both to develop a method to “calibrate” fluorescence (Chla, CDOM) proxies in term of their biogeochemical counterparts over the float life-time, and to evaluate this method through a comparison of multi-float measurements with surface Ocean Color Radiometry remote sensing of the same quantities.

Data and Method

For the BPF, the choice of the measured variables was initially guided both by scientific reasons and by the available technology. The PABIM working group selected 5 core parameters according to four criteria (see D’Ortenzio, 2010) to be measured by the biogeochemical sensors on autonomous profiling floats : The Chlorophyll-a concentration [Chl a] ; The Dissolved Oxygen concentration [DOC] ; The Particulate Organic Carbon [POC] ; The Colored Dissolved Organic Matter [CDOM] and the nutrient Concentration.

Chlorophyll-a is crucial in photosynthesis, absorbing solar energy and allowing its transformation into chemical energy. In the Oceans, the presence of chlorophyll is essentially due to marine phytoplankton which plays a key role in the global climate system. Understanding its spatio-temporal variability by using chlorophyll-a concentration is an important goal of the present day oceanography.

Two methods are implemented on autonomous platforms to estimate chlorophyll-a concentration through independant bio-sensors : the fluorescence-based methods and the radiometric inversion of light measurements. But both methods need to be calibrated for two reasons :

  1. the transformation of the raw fluorescence signal into a so‐called “Chl a equivalent concentration,” as made through the use of the manufacturer’s constant scale factor, provides only a rough indication. Therefore, the calibration of the fluorometer in terms of realistic chlorophyll a concentration, [Chl a] remains to be made on a local basis.
  2. the irradiance inversion depends on the accuracy of the radiometric data which mainly depend on sky conditions, precisely unknown at each profile location.

The proposition of this study is thus to try to find out a help for restoring the entire irradiance profile (which missing parts are due to intermittent cloud) by using the fluorescence profile. Fluorescence can detect a change in chlorophyll a concentration, which impacts the optical properties, and thus the diffuse attenuation coefficient and hence the irradiance profile. Reciprocally, the attenuation coefficient obtained by irradiance and seen as a proxy of [Chl a], can help in calibrating the fluorescence data. The rationale is thus to take advantage of this link that exists between fluorescence and irradiance profiles, at least in case of « clean » waters (no cloud perturbations), between the diffuse attenuation coefficient and the chlorophyll-a concentration.

In 2008, eight PROVBIO floats (Le Reste et al., 2009) with miniaturized radiometers and fluorometers calibrated for chlorophyll and CDOM, were deployed in 4 different oceanic regions (Mediterranean, Iceland Basin, North Pacific and South Pacific) by the LOV-CNRS (PI H. Claustre). They collected data over a time period of about 2 years, by measuring in particular 0–400 m vertical profiles of the downward irradiance at three wavelengths (412, 490, and 555 nm) and of the chlorophyll-a fluorescence.

Results

Thanks to an iterative process which combines irradiance and fluorescence measurements, the fluorescence profiles are locally calibrated to retrieve the chlorophyll a concentration and the irradiance profiles are totally reconstructed (as illustrated figure 2).

Figure 1 : Restored irradiance profiles (heavy lines) superimposed on the initial profiles (dashed lines) for the Mediterranean cast (left) and the North-Atlantic cast (right)

In addition, a validation with regard to the chlorophyll retrieval from the Argo floats data is done thanks to a comparison with the chlorophyll a concentration derived at global scale from space via the data of the (NASA) MODIS‐A sensor. The time series (figure 3) demonstrate for many points a very good agreement in the Mediterranean Sea and in the North Atlantic. The relative [Chl a] variations, as detected by fluorescence or from space, are remarkably coinciding in the North Pacific gyre. In the Southern Pacific gyre, where the lowest [Chl a] values occur, the agreement is rather poor (for the float B07), and more satisfying (for B04) but restricted to only few points.

Figure 2 : Time series of the 8 bio-Argo floats chlorophylle-a concentration derived from the floats measurements (open circles), plotted versus the near-surface[Chl a] values (approximately at the same location and dates) derived from the MODIS‐A radiometric data (square points). Note the large differences in the [Chl a] scales and content, according to the location.

Conclusion

This method is successfully applied to about 600 irradiance and fluorescence profiles measured by bio-Argo floats. Validation of the results in terms of [Chl a] is first made by matchup with satellite (MODIS‐A) chlorophyll (24.3% RMSE, N = 358). A second validation of the method is then obtained by applying it on similar field data acquired from ships, which, in addition to irradiance and fluorescence profiles, include the [Chl a] HPLC determination, used for final verification.

References
  • Boss, E. D. Swift, L. Taylor, et al, 2008: Observations of pigment and particle distributions in the western North Atlantic from an autonomous float and ocean color satellite. Limnology and Oceanography, 53 (5), 2112-2122.
  • Boss, E., M. J. Perry, D. Swift, L. Taylor, P. Brickley, J. R. V. Zaneveld, and S. Riser (2008), Three Years of Ocean Data From a Bio-optical Profiling Float, Eos Trans. AGU, 89(23), doi:10.1029/2008EO230001
  • Claustre et al. (2010), “Bio-optical profiling floats as new observational tools for biogeochemical and ecosystem studies”. Proceedings of the OceanObs09 : Sustained Ocean Observations and Information for Society » Conference, Venice, Italy, 21-25 September, Hall, J., Harrison D.E., Stammer, D., Eds, ESA Pubblication WPP-306, 2010
  • D'Ortenzio, F. and the PABIM group (2010). «White book on Oceanic Autonomous Platforms for biogeochemical studies : instrumentation and measure » -Version 1.3,February 2010 
  • IOCCG (2011). Bio-Optical Sensors on Argo Floats. Claustre, H. (ed.), Reports of the International Ocean-Colour Coordinating Group, No. 11, IOCCG, Darthmouth, Canada.
  • Le Reste, S., André, X., Claustre, H., D'Ortenzio, F. & Poteau, A. (2009). First Succes of ProvBio floats. In : Coriolis letter #5, 6-8.
  • Xing, X., Morel, A., Claustre, H., Antoine, D., D'Ortenzio, F., Poteau, A., Mignot, A. (2011). Combined processing and mutual interpretation of radiometry and fluorimetry from autonomous profiling Bio-Argo Floats. The retrieval of Chlorophyll a, Journal of Geophysical Research, 116, C06020, doi:10.1029/2010JC006899.
  • Xing, X., Morel, A., Claustre, H., D'Ortenzio, F., and A. Poteau (2012). Combined processing and mutual interpretation of radiometry and fluorimetry from autonomous profiling Bio-Argo Floats. II The retrieval of CDOM absorption, Journal of Geophysical Research, 117, C04022, http://dx.doi.org/10.1029/2011JC007632