PANTHYR hyperspectral water reflectance - VEIT

Introduction 

This dataset contains water-leaving radiance reflectance (𝜌𝑤, variable names reflectance and reflectance_nosc) measurements made by an autonomous Pan and Tilt Hyperspectral Radiometer (PANTHYR, Vansteenwegen et al. 2019) installed at site VEIT. Data are provided in NetCDF format with information on processing settings provided in the NetCDF global attributes.  This dataset contains measurements from the first two deployments (Oct. 2019—Oct. 2020 and Nov. 2020—Mar. 2022) that pass quality control, and have bounding calibration information, and ancillary wind speed available. For this site, the use of reflectance with the Similarity Spectrum offset correction (Ruddick et al. 2005, 2006) is recommended.

Methods  

A PANTHYR was deployed at Acqua Alta Oceanographic Tower, Adriatic Sea, VEIT, located at 45.3139°N, 12.5083°E for two deployments Oct. 2019—Oct. 2020 and Nov. 2020—Mar. 2022. 

PANTHYR consists of a pair of TriOS RAMSES radiometers, one for measurement of radiance, and one with a cosine collector for measurement of irradiance, with custom control hard- and software, mounted on a pan and tilt head.  The RAMSES spectral range is about 350—950 nm in 190 channels. The pan and tilt head allows the orientation of each radiometer in a specific direction. Using the standard protocol, a PANTHYR cycle consists of sequential measurements of downwelling irradiance (𝐸𝑑, 6 replicates), downwelling (sky) radiance (𝐿𝑑, 6 replicates), and upwelling radiance (𝐿𝑢, 11 replicates). Three 𝐸𝑑 and 𝐿𝑑 measurements are performed each before and after the 𝐿𝑢 measurements. Measurement cycles are performed every 20 minutes during daytime, at 90, 135, 225, and/or 270 degrees relative azimuth to the sun to minimize air-water interface reflectance (Mobley 1999, Ruddick et al. 2006). Platform pointing conditions are skipped by the definition of an absolute pointing azimuth keep-out zone. Each cycle takes between several minutes (first deployment) and less than a minute (second and later deployments). 

Measurements are converted from digital counts to (ir)radiance using two laboratory instrument characterisations performed by Tartu Observatory (Estonia) before and after each deployment period. Calibration data for a specific scan are obtained from linear interpolation in time between pre-deployment and post-deployment instrument characterisation. The calibrated scan data are linearly interpolated from the instrument specific wavelengths to a common wavelength grid (355—900 nm, every 2.5 nm). Individual calibrated scans are subjected to quality control as in Ruddick et al. (2006), i.e. scans differing > 25% at 550 nm from their neighbouring scans are rejected. For the Ed measurements, this quality control step takes the change in sun zenith angle between the measurements into account.

If sufficient calibrated scans are available in the cycle, i.e. >=5/6 𝐸𝑑, >=5/6 𝐿𝑑, >=9/11 𝐿𝑢, the scans are mean averaged and the standard deviation is computed. The water-leaving radiance reflectance (𝜌𝑤, variable name reflectance_nosc) is then computed according to:

reflectance_nosc = 𝜋/𝐸𝑑 × (𝐿𝑢 - 𝜌𝐹 × 𝐿𝑑)

where 𝐸𝑑, 𝐿𝑢, and 𝐿𝑑 are the mean averaged values, and 𝜌𝐹 the effective Fresnel correction factor as determined from lookup tables provided by Mobley (1999). Ancillary wind speed is obtained fromthe GDAS1 0.25 degree global model 6 hourly nowcast archive, by interpolation of the model grid in time and space to the measurement average time, and site position. The used wind speed and 𝜌𝐹 are provided in the global attributes of each file.

In the present dataset, reflectance data are provided with and without a "nosc" suffix, indicating whether a residual correction for the air-water interface reflectance (Ruddick et al. 2005) is performed. The reflectance without the "nosc" suffix uses the Similarity Spectrum (Ruddick et al. 2006) to estimate a spectrally flat residual air-water interface reflectance error (𝜀) using the 720 and 780 nm combination:

reflectance = 𝜋/𝐸𝑑 × (𝐿𝑢 - 𝜌𝐹 × 𝐿𝑑) - 𝜀

𝜀 = (𝛼 × 𝜌𝑤 780 – 𝜌𝑤 720) / (𝛼 -1),

where 𝛼 is the Similarity Spectrum ratio between the two used wavelengths, i.e. 2.35 for 720:780 nm. The 𝜀 value is provided in the global attributes of each file. For this VEIT dataset, the use of reflectance with Similarity Spectrum correction is recommended.

The reflectance products are further quality controlled using the following criteria:

1) 𝐿𝑑/𝐸𝑑 at 750 nm < 5%, removing non-clear sky conditions

2) Variability (coefficient of variation) of water reflectance at 780 nm < 10%, removing highly variable water conditions

3) Water reflectance > 0 for 350—900 nm, removing spectra with negative reflectance retrievals

4) NIR water reflectance (840—900 nm) is assumed to be decreasing with wavelength, removing potentially contaminated spectra

5) Bright water spectra (average VIS reflectance 400—700 nm > 0.07 or average NIR reflectance 780—950 nm > 0.01) have a local maximum at around 810 nm (805—815 nm) due to the local minimum in pure water absorption, removing potentially contaminated spectra

6) Irradiance measurements in the range 860—885 nm are within 20% of the Gregg and Carder (1990) clear sky model with an aerosol optical depth of 0.1 at normal pressure, removing cloudy, shadowed, or very hazy conditions

Acknowledgements

The installation and maintenance of the PANTHYR at AAOT was carried out within the context of the HYPERNETS project funded by the European Union's Horizon 2020 research and innovation programme (Grant agreement n◦ 775983) and of the HYPERNETS-POP project funded by the European Space Agency (contract n◦ 4000139081/22/I-EF). The skipper and crew of the AAOT and R/V Litus and are also acknowledged.

References

Gregg, W.W. and Carder, K.L., 1990. A simple spectral solar irradiance model for cloudless maritime atmospheres. Limnology and oceanography, 35(8), pp.1657-1675.

Mobley, C.D., 1999. Estimation of the remote-sensing reflectance from above-surface measurements. Applied optics, 38(36), pp.7442-7455.

Ruddick, K., De Cauwer, V. and Van Mol, B., 2005, August. Use of the near infrared similarity reflectance spectrum for the quality control of remote sensing data. In Remote Sensing of the Coastal Oceanic Environment (Vol. 5885, p. 588501). SPIE.

Ruddick, K.G., De Cauwer, V., Park, Y.J. and Moore, G., 2006. Seaborne measurements of near infrared water‐leaving reflectance: The similarity spectrum for turbid waters. Limnology and Oceanography, 51(2), pp.1167-1179.

Vansteenwegen, D., Ruddick, K., Cattrijsse, A., Vanhellemont, Q. and Beck, M., 2019. The pan-and-tilt hyperspectral radiometer system (PANTHYR) for autonomous satellite validation measurements—Prototype design and testing. Remote Sensing, 11(11), p.1360.