Radiances in the MISR Level 1 products processed February 2001 and beyond are based upon a new release of the In-flight Ancillary Radiometric Product (ARP) database. These files will replace any previous ARP data files, and should be used both to process data acquired into the near future, and also when reprocessing earlier data. These new ARP files are:
|T002||Feb 24, 2000||16:41:00||995||T002_F02_0005|
|T003||Jun 12, 2000||4:16:43||2575||T003_F02_0001|
|T004||Aug 29, 2000||14:25:43||3717||T004_F02_0001|
|T005||Nov 1, 2000||21:06:26||4653||T005_F02_0001|
|T006||Dec 19, 2000||19:29:18||5351||T006_F02_0001|
|T007||Mar 7, 2001||01:17:44||6476||T007_F02_0001|
* Filename convention:
000x = version of this Time window file
These files account for instrument degradation with time, in that data from the bi-monthly calibration experiments were used to create the individual time range (T00x) files. The radiometric scale for these files has been established based upon results of the June 11, 2000 vicarious calibration experiment. Data from this experiment was used to re-calibrate the in-flight radiometric standard, i.e., the Blue filtered, High Quantum Efficiency (HQE) photodiode. Although temporally stable, the photodiode response had previously been unverified. With this photodiodein-flight calibration, radiances produced from MISR have been increased by 9%, and now agree with radiances measured by AirMISR and the vicarious calibration in-situ measurements.
Although continued validation of our approach is required, we believe that typical MISR data products are accurate in an absolute sense to the 4 percent level at 1.0 equivalent reflectance level (one sigma confidence level). At the lower end of the cameras dynamic range response (0.001 in equivalent reflectance), the uncertainty increases to 10% due to photon noise limitations, and possible sensor non-linearities. Known exceptions to these accuracy levels are the An Red and NIR channels processed using the T002 through T004 files. These channels have an absolute radiometric uncertainty of 6% at full scale. This is based upon an anomalous change in response, as compared to a smooth degradation trend. (These gain coefficient values will be revised in the future.) Another exception to this accuracy statement is with respect to early mission data (February - March, 2000). During this time period the MISR radiometric response was rapidly changing, due to an initial instrument on-orbit settling effect. (This change has been noticed in most other on-orbit sensors, and is generally attributed to the browning of the optics in the space environment. This browning typically stabilizes after a few months.) It is for this reason that use of ARP T002 may result in radiometric errors as large as 10%, for data acquired in the February/ March time era. One proposal to reduce uncertainties for this mission-start time period is to report radiometric degradation versus time on the MISR calibration web site. The data users for these early products could then adjust the radiometric values based upon the date of data acquisition, and the web-published instrument degradation with time model.
To assess the radiance change with ARP revision, a summary of the channel-average gain coefficient versus ARP data file is posted to the MISR web site, http://www-misr.jpl.nasa.gov/mission/valwork/cal_reports/arp.html. These data will additionaly allow an investigator to produce an approximate Level 1 data product, based upon the ARP updates, without having to reorder data. The procedure is good to about the 1% level.