Spatio-temporal variability of the wet component of the troposphere – Application to satellite altimetry

Abstract

Due to the presence of water vapour and cloud liquid water in the atmosphere, the wet component of the troposphere is responsible for a delay in the propagation of the altimeter signals, the Wet Path Delay (WPD). The high space–time variability of the water vapour distribution makes the modelling of WPD difficult, its effect still being one of the main error sources in satellite altimetry applications, e.g. in the estimation of Mean Sea Level (MSL). The understanding and the quantification of the WPD variability on various spatial and temporal scales are the main purposes of this study, in view to improve the MSL error budget. The dominant timescales of WPD variability and its correlation with Sea Level Anomaly (SLA) are examined. In these analyses, the atmospheric reanalysis ERA-Interim model from the European Centre for Medium-Range Weather Forecasts (ECMWF) is used to derive a global dataset of daily grids of WPD, spanning a 28-year period from January 1988 to December 2015. The Seasonal-Trend decomposition procedure based on Loess (STL) is used to extract precise WPD annual and interannual signals. Linear trends have been derived from the interannual time series and the contribution of each STL component was mapped globally, allowing the understanding of the WPD variability in spatial terms. The correlation between SLA and WPD is mapped and decomposed into seasons using monthly mean grids, for a period of 21-years, from January 1993 to December 2013.Aiming at inspecting the sensitivity of the results to the used data set, the WPD temporal analysis is extended to the data set provided by the Special Sensor Microwave Imager (SSM/I) and SSM/I Sounder (SSM/IS) Sensors. The WPD from SSM/I(S) is compared against those from the ERA-Interim and from the National Centers for Environmental Prediction (NCEP).Results show that climate phenomena, especially the El Niño Southern Oscillation (ENSO) are the cause for this high variability, since they affect the water vapour and temperature. The observed trends from ERA-Interim, computed globally and over ocean regions only, allow concluding that WPD is increasing with time by approximately 0.1 mm per year, and the maximum trends are observed for the Pacific North and Indian Oceans. High correlation between WPD and SLA is found over the western tropical Pacific.The comparison between WPD from SSM/I(S) and from ERA-Interim and NCEP, allows concluding that the trends computed using only the SSM/I(S) measurement points are substantially larger.