February 9, 2006
<NASA TSDIS Announcement>
TRMM Science Users:
An error has been found in the PR 3A25 6A product that affects the storm height statistics for the high resolution 0.5 x 0.5 degree grid. This is an error in the TSDIS Toolkit, not the algorithm.
The storm height mean and standard deviation fields for the high resolution grid consist of three grids (those arrays should have 3 indices). The first index is for stratiform rain, the second for convective and the third for total rain. The last index for total rain was not being written into the product by the toolkit. Those fields only have 2 indices, contrary to the file specifications. The Toolkit code structure contains 3 indices for these fields, however,the last one is not populated with data when reading the 3A25 product.
This error only affects the high resolution data. The low resolution grid (5.0 x 5.0 degrees) is written correctly with all three fields.
Total = stratiform + convective + other rain.
The other rain classification makes up about 2.5% of all rain. In addition, the difference between total and (stratiform + convective) is very small in terms of storm height means. A test product correcting the error was generated and additional data can be provided upon request.
It has been decided not to change the Toolkit or the 3A25 algorithm since we are so far along in this processing cycle. This decision is approved by the PR Team Leader and by the TRMM Project Scientists in Japan and U.S. The documentation for 3A25 will be updated to have only 2 indices (stratiform and convective) for the high resolution storm height fields.
Thank you.
TSDIS