The anomaly was detected looking at time series of N2O values of MIPAS reprocessed L2 data.
Looking at the figure below the anomaly appears as excessive N2O concentration at certain time periods.
The anomaly affects L2 MIPAS data processed OFL (with software version 4.61, 4.62) and available at D-PAC archive.
The anomaly is observed for the following mission interval:
from March 7 18:00 to March 8 19:00 2003
from March 13 18:00 to March 15 4:00 2003
The anomaly is not latitude dependent and is observed also at high altitude (up to 30km).
The anomaly affects all the retrieved species with different feature:
An excessive retrieved VMR for N2O, CH4 and O3
A deficient retrieved VMR for H2O, NO2 and HNO3
An excessive chi-square value is detected in correspondence to this anomaly for all the retrieved species, in particular for N2O, p-T and O3.
Most of the L2 products affected by this anomaly have product error flag set to 1 in the MPH. This flag came from the corresponding L1 products used as input.
The used L1 products show many corrupted observations in the band A, AB and B, while bands C and D are less affected by corruption
The cause of the anomaly is the usage of wrong ADFs. The problem was that in the ADF database for the re-processing of FR MIPAS mission there was a gap in the ADF validity, in particular for the time period of the N2O anomaly the L1 processor did not find any available ADF, then it used a very old set (taken from commissioning phase!) with a resulting wrong spectra calibration. This finally causes the anomalous retrieval.
The affected products were removed from the server and re-processed with correct AUX files.