Interannual variability of upper tropospheric and lower stratospheric (UTLS) region over Ganges–Brahmaputra–Meghna basin based on COSMIC GNSS RO data

Abstract

Abstract. Poor reliability of radiosonde observational networks across South Asia imposes serious challenges in understanding climate variability and thermodynamic structure of the upper-tropospheric and lower-stratospheric (UTLS) region. The Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) mission launched in April 2006 have overcome many observational limitations that are inherent in conventional atmospheric sounding instruments. This study investigated the interannual variability of the UTLS region over the Ganges–Brahmaputra–Meghna (GBM) basin based on COSMIC radio occultation (RO) data from August 2006 to December 2013. Detailed comparisons were also made with various different radiosonde types and Numerical Weather Prediction (NWP) products. The results indicated that Indian Meteorological Department (IMD) radiosondes performed poorly despite upgrading to newer techniques. ShangE (of China) sonde showed the best agreement with COSMIC RO data with a mean temperature difference of −0.06 °C and a standard deviation of 1.44 °C while the older version (ShangM) indicated a cold bias of 0.61 °C in the UTLS region. The inter-annual variability of temperature in the UTLS region based on COSMIC RO data indicated a clear pattern of El Niño Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) while the stratospheric temperature anomalies reflected all three major Sudden Stratospheric Warming (SSW) events since 2007. The mean tropopause temperature varied from −70 to −80 °C, with an average height of about 15.5 to 16.3 km from winter to summer, indicating a pronounced annual cycle. The annual amplitudes of tropopause were found to be in the order of 0–6 °C and 0–1.5 km from tropical south to subtropical north. The anomalies of tropopause temperature and height exhibited the patterns of ENSO and IOD exceptionally well with a correlation of 0.65 and −0.52, respectively. The temperature data from Modern-Era Retrospective Analysis for Research Application (MERRA) agreed very well with COSMIC RO data.