1- Introduction
The Community Microwave Emission Modelling Platform (CMEM) has been developed by the European Centre for Medium-Range Weather Forecasts (ECMWF) as the forward operator for low frequency passive microwave brightness temperatures from 1GHz to 20 GHz of the surface. Up to date CMEM documentation is provided in de Rosnay et al 2019.
It is a new highly modular code providing I/O interfaces for the Numerical Weather Prediction Community. CMEM's physics is based on a range of State-Of-The-Art parametrisations, including those used in the L-Band Microwave Emission of the Biosphere (LMEB, Wigneron et al., 2007) and Land Surface Microwave Emission Model (LSMEM, Drusch et al., 2007). So, CMEM modularity allows considering different parametrisations of the soil dielectric constant as well as different soil approaches (either coherent of incoherent), different effective temperature, roughness, vegetation and atmospheric contribution opacity models.
The SMOS satellite, launched on 2 November 2009, is the first instrument to provide global fields of L-band brightness temperature. SMOS brightness temperatures is used at the European Centre for Medium-Range Weather Forecasts (ECMWF) to investigate its use to analyse soil moisture through the Surface Data Assimilation System and to monitor ocean salinity. This is expected to improve the accuracy of initial conditions of the Numerical Weather Prediction (NWP) model. In turn, NWP products are of great importance for space agencies in order to derive the Level2 soil moisture and ocean salinity products. CMEM is used at ECMWF to monitor and assimilate the SMOS brightness temperature in the Integrated Forecast System (IFS).
In the latest version (ver.6.0), it is possible to simulate brightness temperature up to 100 GHz by option.
2- Implementation
The Community Microwave Emission Modelling Platform was developed and is maintained at ECMWF in the context of an ESA contract for the SMOS mission.
Operational numerical weather forecast systems are widely used to evaluate and analyse new types of satellite observations. Numerical weather prediction (NWP) centres are prime customers as observations are used in the analyses to derive Level2 retrieved geophysical parameters (eg soil moisture or ocean salinity for SMOS) from the observed brightness temperatures or radiances.
Before the SMOS launch, forecast systems can be used in the product definition phase. Based on modelled atmospheric and land state variables, the effect of different parameterizations and auxiliary data sets on the retrieval can be analysed. In addition, the errors introduced through temporal collocation mismatches between the auxiliary data sets and the actual observations can be quantified.
After the SMOS launch, when real SMOS observations are available, monitoring, i.e. comparison between the modelled equivalent of the observation and the observation itself, makes a significant contribution to the calibration / validation activities. Any systematic error or spikes, become visible and can be reported to ESA and the other calibration and validation teams without significant delays.
In this context, CMEM implementation strategy included the development and implementation of the CMEM forward model for SMOS level 1 data at ECMWF for quality monitoring and the development of the assimilation scheme for SMOS level 1c brightness temperature in ECMWF's global NWP system.
For the atmospheric radiative transfer calculations, the RTTOV [Radiative Transfer model for Television Infrared Orbiting Satellite (TIROS) Operational Vertical Sounder (TOVS)] software package has been developed as a community model. It is updated and maintained by the UK Met Office under the framework of EUMETSAT's NWP Satellite Application Facility (SAF). Although CMEM has been designed for frequencies below 20 GHz, its modular structure allows upgrades to higher frequencies and it is envisaged to interface CMEM into the RTTOV software package.
3- Platform structure
CMEM is coded in Fortran 2003. It is a highly modular software package providing I/O interfaces for the Numerical Weather Prediction Community.
The individual parameterizations provided in the CMEM Platform are detailed in the SMOS ATBD (2007), in L-MEB (developed and maintained by INRA/EPHYSE in France, Wigneron et al., 2007, Pellarin et al., 2003) and LSMEM (Drusch et al., 2001) (see section on the platform modelsn physics below).
Based on these parameterizations, CMEM was specifically designed to be highly modular in terms of both physics and input/output interface. For the user, this allows choosing between different options in the simulation definition without requiring any change on the code.
The CMEM code is organized as described in the Figures. The main program is cmem_top.F90.
Radiative transfer computation of CMEM is modular. It is structured in four modules for soil, vegetation, snow and atmosphere. The Platform choice of different models physics is detailed next section.
4- Platform models physics
CMEM was originally developed at ECMWF for Numerical Weather Prediction applications, namely as a the forward operator for the assimilation of low frequency passive microwave observations (Holmes et al., 2008). CMEM has been extended to be a platform with modular Input/Output that include a range a physical parameterizations, some of which are also used in the Land Surface Microwave Emission Model (LSMEM) and the L-band Microwave Emission of the Biosphere (L-MEB). It comprises four modules for computing the contributions from soil, vegetation, snow and the atmosphere. The code is designed to be highly modular and for each microwave modeling component, a choice of several parameterizations are considered. More information in de Rosnay et al 2019.
References:
de Rosnay, P., J. Muñoz-Sabater, C. Albergel, L. Isaksen, S. English, M. Drusch, J.-P. Wigneron: "SMOS brightness temperature forward modelling and long term monitoring at ECMWF", 237 (Feb 2020) https://doi.org/10.1016/j.rse.2019.111424