Background

The performance of the SeaWiFS nLw have demonstrated remarkable consistency for most of the mission to date. Recently, analyses of the interannual repeatability have shown measureable deviations from the mission trends. Starting in late 2005, the global nLw averages have deviated from the previous results by typically 3% to 5%.

One possible cause for the recent change is the orbit node drift. The SeaWiFS orbit was originally specified with a noon descending node crossing time, within 15 minutes of local noon. The crossing time remained within these limits during the nominal (first five years) mission without the need for orbit maintenance. At the end of this period, the crossing time was drifting into the afternoon, and has continued to drift at an increasing rate, with the current crossing time being about 12:56 PM local time. The node crossing time is displayed on the SeaWiFS Mission Operations web site for the last 30 days, last year, and the entire mission (note that the plots show the node crossing in degrees, where 1 degree equals 4 minutes):

http://seawifs.gsfc.nasa.gov/plankton2/d3/userweb/orb/plots/node.gif

One result of the orbit node drift is an equivalent drift in the Sun angle relative to the orbit plane and the spacecraft (variously referred to as the Sun roll, or beta, angle). The Sun roll has an annual cycle that results from the Earth's orbit eccentricity and polar inclination and the SeaWiFS orbit inclination, and a long-term drift from the orbit node, as shown in the long-term trend plot on the Mission Ops site:

http://seawifs.gsfc.nasa.gov/plankton2/d3/userweb/tlm/tlm_analysis/summary_plots/median_zoom_in/SAC_sun_roll.gif

The drift in the Sun angle has caused a change in the thermal environment of the spacecraft. For calibration purposes, the most important temperature is that of the instrument focal planes. The focal plane temperatures have shown an annual cycle that has been driven primarily by the Sun-Earth distance; this distance is a minimum in early January, and the temperatures have shown a peak at that time of year. However, as the orbit node and Sun angle have drifted, the focal plane temperatures have deviated somewhat from this behavior. The long-term trend for the band 7/8 focal plane temperature is shown from the Mission Ops site:

http://seawifs.gsfc.nasa.gov/plankton2/d3/userweb/tlm/tlm_analysis/summary_plots/median_zoom_in/INST_band_78_temp.gif

This plot clearly shows the annual peak in January over the mission. As the Sun roll angle changes, the maxima and minima in annual temperature cycle decreases. This change in the focal plane temperature cycle is being analyzed as a possible cause of the deviation in the nLw trends.

Current Analysis

The operational SeaWiFS Level-1B algorithm corrects for changes in the radiometric response of the instrument due to variations in the focal plane temperatures. The algorithm assumes that the focal plane temperature corrections are constant over time.

The time series of residuals of the exponential fits for band 8 are shown in the figure below. The residuals are correlated with the focal plane temperatures shown above. The fit residuals show a change in the mission-long behavior around 1 January 2006 -- there is a jump in the annual cycle of the residuals at this time.

The residuals of the fits, when plotted against the focal plane temperatures, allow the focal plane temperature dependence of the detector response to be evaluated, as shown in the figure below. The mission-long trend is shown in blue, while the trend since 1 January 2006 is shown in red. The Cal/Val team has used the fits to the residuals as functions of temperature to compute a revised set of temperature corrections for SeaWiFS. The mission-long trend provides a set of corrections from the launch through 1 January 2006, while the emergent trend provides a set of corrections from 1 January 2006 onwards.

The lunar calibration time series, processed with the revised focal plane temperature corrections, are shown here.

Return to SeaWiFS Lunar Calibration Time Series.

For further information contact Gene Eplee.

Last modified: Wed Jun 6 16:17:06 EDT 2007