The effects of the revised focal plane temperature corrections can be observed over a single orbit. As part of the reprocessing #4 analysis effort, we have examined these effects for a single orbit on May 1, 1999 for bands 7 and 8.
The focal plane temperature for the band 7/8 focal plane for a typical orbit on May 1, 1999 is shown in Figure 1 as a function of the latitude of the center pixel of the scan line. The digitization of the temperature is 0.2 C. The behavior of the other four focal planes is similar. During the back orbit, the detectors on the focal plane are turned off and the focal plane heater is turn on. As the spacecraft crosses over the north pole, the focal plane heater is turned off and the detectors are turned on. Typically, the focal plane heater keeps the focal plane at a higher temperature than does the normal operation of the detectors. As shown in Figure 1, the focal plane temperature falls from the temperture of the heater (~16.8 C on this orbit) to a more typical temperature of ~14.1 C within about the first 5 degrees of latitude. The temperature continues to fall to ~13.7 over the next 15 degrees of latitude. This temperture, which is the nominal operating temperature of the focal plane for this orbit, through 55 degrees of latitide. As the spacecraft approaches the equator, solar heating of raises the tempreature to ~13.9 C, and then to ~14.1 C. For most of the orbit, from 75 degrees north to 60 degrees south, the focal plane temperature is stable to within two counts (0.4 C).
The focal plane temperature corrections for bands 7 (blue) and 8 (red) are shown for over the orbit as functions of latitude in Figure 2. The prelaunch corrections are plotted as dashed lines and the revised corrections are plotted as solid lines. The variation in the prelaunch corrections over the orbit is essentially unity. Over the temperature range of 14-17 C, the revised temperature correction for band 8 is ~0.05% per count and the correction for band 7 is ~1/3 smaller. The change in the temperature correction for band 8 is ~0.1% over the course of the orbit. Since the band 8 radiance is used in the atmospheric correction algorithm to determine the aerosol abundance, the decrease in radiance in the southern hemisphere may affect the retrieved aerosol abundance between reprocessing #3 and reprocessing #4 for a single pixel.
The combined temperature and time corrections affect the total radiances. In this analysis, single time corrections for bands 7 and 8 have been computed for May 1, 1999 and use to scale the focal plane temperature correction over the orbit. The band 7/ band 8 ratio, normalized by the vicarious gain of band 7, is used to determine the aerosol type in the atmospheric correction algorithm. Figure 3 shows the ratio of (gain 7*band 7/band 8) for reprocessing #4 to reprocessing #3 as a function of latitude. The plot shows that for the southern hemisphere the change in the band 7/band 8 ratio between reprocessing #3 and reprocessing #4 may cause a change in the retrieved aerosol type. The changes in aerosol retrievals which result from the reprocessing #4 differences shown in Fig. 2 and 3 will propagate through to changes in the retrieved water-leaving radiances and chlorophyll concentrations.
The effects of the revised focal plane temperature corrections and resulting exponential time correction are observed on a monthly basis in the lunar calibration data. As part of the reprocessing #4 analysis effort, we have also examined these effects on daily data for bands 7 and 8, where the temperature correction revisions are the most significant.
Figure 4 shows the mean daily focal plane temperature for the band 7/8 focal plane over the course of the mission. The tempertures range from 14 C to 20 C on an annual cycle as the Earth-Sun distance changes. Figure 5 shows the prelaunch (red) and revised (blue) band 8 focal plane temperture corrections corrresponding to the mean daily temperatures. Figure 6 shows the corrections for band 7. These figures show that the prelaunch corrections for bands 7 and 8 are essentially unity, while the revised corrections change the radiance in band 8 by ~1% and the radiance in band 7 by ~0.3% over the course of a year.
The combined temperature and time correction to the radiances for band 8 over the course of the mission are shown in Figure 7. The prelaunch temperature correction and quadratic time correction (reprocessing #3) are shown in red and the revised temperature correction and exponential time correction (reprocessing #4) are shown in blue. Figure 8 shows the corrections for band 7. For both bands, the effect of the revised temperature correction on combined radiance correction is apparent. Figure 9 shows the ratio of the reprocessing #4 combined correction to the reprocessing #3 combined correction for band 7 (blue) and band 8 (red). This plot shows the change in the total radiances in bands 7 and 8 at any point in the mission arising from the changes in the focal plane temperature and time corrections.
Bands 7 and 8 are used for atmospheric correction of the ocean data. Band 8 is used to determine the aerosol abundance and the band 7/band 8 ratio, normalized by the vicarious gain of band 7, is used to determine the aerosol type or model. Figure 10 shows the ratio of (gain 7*band 7/band 8) for reprocessing #4 to reprocessing #3. The plots of the band 8 ratios in Figure 9 and the band 7/band 8 ratios in Figure 10 show that at any point during the mission, changes in the band 8 radiance and in the band 7/band 8 ratio from reprocessing #3 to reprocessing #4 may result in changes in the retrieved aerosol abundance and type for a given pixel. These changes in the aerosol retrievals will propagate through to changes in the retrieved water-leaving radiances and chlorophyll concentrations.
