The OCTS was calibrated prelauch using an integrating sphere owned by the NEC Corporation and operated at the Japanese National Research Laboratory of Meterology. This calibration provied the preflight commanded instrument gain settings for each band and the gain slope and offsets for the detector normalization factors. OCTS has four instrument gain settings: (1) Ocean Normal, (2) Ocean High, (3) Land Normal and (4) Land High. A tilt reflectance normalization correction was also derived to account for the BRDF of the mirror at the different instrument tilt states.
The OBPG has created an HDF caltable that contains these prelaunch calibration coefficients.
It was determined that the tilt reflectance normalization coreection factors introduced an obvious discontinuity in the OCTS time series. As a result of this discovery, the OBPG processing sets these tilt reflectance factors to unity. The figure below shows the impact of the removal of this tilt normalization factor on the retrieved water-leaving radiance at 412nm. The offset between the gray (with tilt reflectance correction) and black lines is the result of a revised calibration. The shift in the gray line evident during the Nadir Tilt operations period is the result of the apparent erroneous tilt reflectance correction.
Of the four instrument gain settings, only the "Ocean Normal" set are deemed usable by the OBPG. Discontinuities in the data where the commanded gain changed were not easily reconcillable and since the majority of the data were collected under the "Ocean Normal" gain, the other gain sets have the 'NAVFAIL' flag set to eliminate these from L3 binning.
It was apparent from early mission long, global processings for OCTS that the instrument did experience some temporal changes in system response (see above figure). However, no on-board or stable external calibration source is available for OCTS. The instrument did have internal incandescent lamps, but these proved insufficient for calibration purposes. The OBPG was able to estimate a temporal change in OCTS system response for the two NIR bands (765 and 865nm) using the inverse calibration technique with the assumptions of negligible water-leaving radiance and a fixed maritime aerosol. A similar method could have been applied to the visible bands using a clear water radiance model to estimate the ocean contribution. However, as OCTS was only operational for less than 9 months and during those months the onset of a significant El Nino occured, it was deemed impractical to do so. The caltable was modified to include the temporal change estimated for the NIR bands, but the visible bands were left unchanged.
The OBPG derived vicarious calibration coefficients in addition to the prelauch gains by using a clear-water radiance model based on a climatological (in situ based) chlorophyll time-series at the Bermuda Atlantic Time Series (BATS) and Hawaiian Ocean Time-series (HOT) sites. MOBY data were availble for the OCTS time period, but these were from a pre-operational deployment of the buoy. The OBPG has confirmed that gains derived from the model based approach are consistent with those derived from the MOBY.
