Oxford-RAL IASI Group Meetings

Meeting#24 (25Apr18 @ Oxford)

Oxford: Elisa Carboni, Anu Dudhia, Lucy Ventress, Don Grainger

RAL: Brian Kerridge, Diane Knappett, Pasquale Sellito, Richard Siddans

Leicester: Mike Cartwright, Dave Moore

CH4 (DK) [Prev] [Next]

Trends in CH4 2007-2012, but need to add offset to cope with May 2013 offset (jump in May 2010 doesn't seem to affect CH4). Not much trend evident 2007-2012. Scale factor for assumed 13CH4 - also no evident trend. Surface temperatures show ENSO signature but negative trend over China (?), and change in emissivity over Australia. 13CH4 shows clear latitudinal structure, may be interference of stratosphere or H2O.

SO2 (EC) [Prev] [Next]

SO2 linear retrieval using just v3 band shows maximum over Himalayas in monsoon season - either more SO2 or higher SO2. But doesn't appear in LV's linear retrieval using whole spectrum and just perturbations up to 15km.

H2SO4 (EC) [Prev]

Retrieve H2SO4 aerosol, based on ash retrieval. MPhys student project finished. Linear retrieval shows Calbuco signature.

IMS (RS) [Prev] [Next]

T,q,emiss data processed 2017 data as well now, including cloudy scenes. Also retrieve cloud fraction and height.

Recent RTTOV update v12.2 - adds CAMS internal aerosol models, better cloud parameterisation using effective radius.

CO (DM) [Prev]

Looking at Canadian fires of 2017, compared with FORLI. Big difference between TOMCAT and FORLI a priori, especially in stratosphere. Try merging single a priori profile for ULIRS CO retrieval. Also tried using RAL IMS as starting guess rather than ECMWF - increases CO column over Indonesia.

OCS (MC) [Prev] [Next]

Running CO simulations using ULIRS, learning how to use the code and diagnostics. Eventually move to OCS.

IASI 2010 discontinuity (LV)

2010 data jump issue. EUMETSAT help desk sent some documents. File format changed but can't find any more information. Some processing was apparently changed. Not caused by switching from BEAT to read_l1c to read L1C data.

NH3 (LV) [Prev] [Next]

Month's worth of GEOS-Chem Apr 2014 over India, compared with linear retrieval. Generally reproduces large scale features. Not yet run MORSE iterative retrieval. Would like RAL IMS data for 2014.

Scaled Linear Retrieval (AD) [Prev] [Next]

Linear retrieval based on joint retrieval of target molecule with singular vectors to remove background spectrum, also a gain matrix scaling factor derived from fitted singular vector coefficients. Working on getting this into simple IDL procedure for processing complete days. NH3 results for Russian fires, two orbits, show good spatial coherence.