Deep SUNA Manual
For SUNA running firmware version 2.4 or later
3. The SUNA Sensor
throughput using its photo-spectrometer. The difference between this measurement and
a prior baseline reference measurement of pure water constitutes an absorption
spectrum.
Absorbance characteristics of natural water components are provided in the sensor
calibration file. The Beer-Lambert law for multiple absorbers establishes the relationship
between the total measured absorbance and the concentrations of individual
components. Based on this relationship, the sensor obtains a best estimate for the
nitrate concentration using multi-variable linear regression.
The approach described above was initially developed at MBARI (cf. Kenneth S.
Johnson, Luke J. Coletti, In situ ultraviolet spectrophotometry for high resolution and
long-term monitoring of nitrate, bromide and bisulfide in the ocean, Deep-Sea Research
I 49 (2002) 1291–1305) and the technology then transferred to Satlantic.
3.3.2 Nitrate Concentration
Nitrate processing uses the 217–240 nm wavelength interval, which contains
approximately 35 spectrometer channels. For each channel, the absorbance is
calculated, and decomposed into individual absorbers using the MBARI method.
The precision of the nitrate concentration depends on the number of absorbers into
which the measured absorbance is decomposed. Thus, in freshwater deployments, the
number of concentrations to be fitted should be set to 1.
High absorbance conditions introduce inaccuracies into the nitrate concentrations.
Therefore, channels with an absorbance greater than 1.3 are excluded from processing.
If less than about 10 channels remain, the sensor is unable to determine a nitrate
concentration, and the measurement is no longer valid (out-of-bounds). Users can
overturn the standard setting and increase the absorbance cutoff, obtaining reduced
accuracy nitrate concentrations at higher absorbances. There is, however, a limit at
around 2.5 absorbance units, when nitrate concentrations can no longer be determined.
3.3.3 Interferences and Mitigation
The quality of the nitrate measurements can be impacted in a number of ways. This
impact has been quantified (see section 3.2.3 Performance Specifications) for some
significant interfering influences. Here, interferences are explained, and mitigation
options are explored.
Sample temperature:
Seawater is known to have a temperature-dependent absorption.
If this effect is not taken into account, a bias and/or imprecision are introduced to the
reported nitrate concentration.
This effect can be mitigated by providing sample temperature and salinity to the nitrate
calculation, either in real-time (supported in APF mode) or in SUNACom post-
processing (collection of spectra and accompanying temperature and salinity data is
Copyright © 2013, Satlantic LP, All rights reserved
16
