On the quality of re-analysis and merged microwave spaceborne soil moisture products: global trends and ability to represent in situ measurements

The importance of soil moisture (SM) in the global climate system has recently been underlined by the Global Climate Observing System (GCOS) by endorsing soil moisture as an Essential Climate Variable. It is a crucial variable for numerical weather and climate prediction and plays a key role in hydrological processes. A good representation of soil moisture conditions can therefore help to improve forecasting of precipitation, droughts and floods, climate projections and predictions. The first SM analysis system used for operational Numerical Weather Prediction was implemented by ECMWF in 1994 to prevent the Land Surface Model from drifting to dry conditions in summer. Ever since then, major upgrades have been implemented in the land surface modelling and analysis systems of the deterministic version of the Integrated Forecasting System (IFS) used operationally at ECMWF. ECMWF also develops Re-analyses such as ERA-Interim. They use a fixed IFS version, for the main component of the atmospheric model and data assimilation. However pace at which they can be re-run, hence benefiting from the most recent model and data assimilation advances, is of the order of 5 to 10 years. To re-actualize only the land surface component of the ERA-Interim reanalysis, ECMWF recently developed ERA-Interim/Land simulation where the ERA-Interim near-surface meteorology is used as forcing term to produce a new land surface model trajectory. In this context, this study proposes an evaluation of various soil moisture products, (i) the above mentioned ERA-Interim/Land from ECMWF, (ii) the Modern-Era Retrospective Analysis for Research and Applications (MERRA) generated by the NASA Global Modelling and Assimilation Office (GMAO) and (iii) the Essential Climate Variable Soil Moisture dataset generated within the Climate Change Initiative (CCI) project using active and passive microwave spaceborne instruments. The different SM data set are first evaluated over the 2007-2010 period using ground based measurements to assess their ability to match the observations. The 1988-2010 period is then analysed for monotonous increasing or decreasing trends with the non-parametric Mann-Kendall test and Sen's method for slope estimates.