MODAPS PCR 10-007
| PCR Number | 10-007 |
|---|---|
| Date | 2010-06-15 |
| Initiator | James Kuyper |
| Abstract | Update PGE01 to extract the correct EulerAngleOrder from the attitude file's metadata, and then call the SDP Toolkit routine PGE_CSC_EulerToQuat() to convert from Euler angles to a corresponding attitude quaternion. |
| Problem Statement | In normal operations, PGE01 uses SDP Toolkit functions to get attitude quaternions from GDAAC attitude files. These quaternions are used solely in the computation of the solar "elevation" angle. That angle is calculated solely to determine thermal stress corrections to the matrix describing the orientation of the MODIS instrument relative to the satellite. When attitude data entrained in the Terra L0 files is used instead of the SDP Toolkit attitude files, it doesn't come with attitude quaternions. Because of difficulties with obtaining the correct Euler angle order, the code was simplified by using a fixed attitude quaternion. This simplification was deemed acceptable, partly because entrained attitude data were not being used in normal production. However, it was also based upon a misunderstanding - that the fixed attitude quaternion being used corresponded to a nadir-pointing orientation of the MODIS instrument, and as a result would only cause very small errors in most cases; in reality, errors of several hundred meters have been observed. With the establishment of the LANCE-MODIS system, the use of entrained data has become more important, and the differences between LANCE-MODIS and normal MODIS processing are just barely large enough to be a problem. |
| Description of Change | Extract the correct EulerAngleOrder from the attitude file's metadata, and then call the SDP Toolkit routine PGE_CSC_EulerToQuat() to convert from Euler angles to a corresponding attitude quaternion. These routines are already linked into the program for internal use by the SDP Toolkit. This was tried in 2004, but with a fixed (and apparently incorrect) EulerAngleOrder. The problem could have been resolved, but the attempt was halted for lack of time. |
| Products Affected | MOD03 |
| Software Affected | PGE01 |
| Processing String to Receive the Change | While this change affects only Terra LANCE-MODIS processing, for the sake of consistency the same new version of the PGE should be installed for Aqua as well as Terra, and for normal operations as well as LANCE-MODIS. |
| Downstream Product Effects | All other MODIS products are downstream of PGE01; and all of them will be affected, directly or indirectly, by the slightly improved accuracy of the geolocation using PGE01 on LANCE-MODIS. It's not as accurate as normal production, but it's substantially better than PGE01 5.0.40 on LANCE-MODIS. |
| Data Granules to be Used for Science Testing | 2010-059 |
| Justification | The difference between LANCE-MODIS products and the more accurate ones from MODAPS dropped from several hundred meters to just a few meters. |
| Effective Date for Implementation of Change | |
| Status | Bill Ridgway approved for Atmospheres, Sadashiva Devadiga approved for Land,Michael King approved on 07/19/2010. |
