This article presents coupled data assimilation (DA) activities at the European Centre for Medium‐Range Weather Forecasts (ECMWF). Coupled DA is an essential component of the ECMWF Earth‐system strategy. It aims at providing consistent initial conditions to the coupled atmosphere, land, and ocean forecast model. The article introduces the different DA systems and observing systems for each Earth‐system component. It discusses challenges related to observation consistency, availability, and sustainability across the components. It gives a review of coupling methodologies and presents coupling methods in development at the ECMWF. The current ECMWF system relies on weakly coupled DA approaches between land and atmosphere, between wave and atmosphere, and between ocean and atmosphere. Research on coupled DA has been centred on outer loop coupling developments and evaluation, focused on ocean–atmosphere coupling for reanalysis and for numerical weather prediction. The latest configuration of the ocean–atmosphere outer loop coupling combines weak and outer loop coupling methods. The article discusses the challenges in assimilating surface‐sensitive observations and it presents opportunities that coupled DA offers to enhance the exploitation of interface observations. It presents recent developments on land–atmosphere forward operator coupling and multilayer snow‐emission modelling and shows the potential for coupled radiative transfer modelling at low‐frequency passive microwave. The article discusses ongoing developments of in‐house sea‐surface temperature analysis based on coupled skin temperature assimilation. It presents preliminary results that demonstrate meaningful upper ocean temperature increments when assimilating skin temperature information from four‐dimensional variational assimilation. Coupled assimilation results are illustrated and discussed in the context of several applications related to reanalysis, scatterometer impact on coupled ocean–atmosphere, tropical cyclone, impact of weakly coupled sea‐ice–atmosphere data assimilation as well as land surface coupled DA case studies. These examples illustrate benefits and challenges of coupled DA developments. The article discusses future plans of coupled DA developments in support of ECMWF's Earth‐system 2021–2030 strategy.

LDAS-Monde is a global offline land data assimilation system (LDAS) that jointly assimilates satellite-derived observations of surface soil moisture (SSM) and leaf area index (LAI) into the ISBA (Interaction between Soil Biosphere and Atmosphere) land surface model (LSM). This study demonstrates that LDAS-Monde is able to detect, monitor and forecast the impact of extreme weather on land surface states. Firstly, LDAS-Monde is run globally at 0.25^∘ spatial resolution over 2010–2018. It is forced by the state-of-the-art ERA5 reanalysis (LDAS_ERA5) from the European Centre for Medium Range Weather Forecasts (ECMWF). The behaviour of the assimilation system is evaluated by comparing the analysis with the assimilated observations. Then the land surface variables (LSVs) are validated with independent satellite datasets of evapotranspiration, gross primary production, sun-induced fluorescence and snow cover. Furthermore, in situ measurements of SSM, evapotranspiration and river discharge are employed for the validation. Secondly, the global analysis is used to (i) detect regions exposed to extreme weather such as droughts and heatwave events and (ii) address specific monitoring and forecasting requirements of LSVs for those regions. This is performed by computing anomalies of the land surface states. They display strong negative values for LAI and SSM in 2018 for two regions: north-western Europe and the Murray–Darling basin in south-eastern Australia. For those regions, LDAS-Monde is forced with the ECMWF Integrated Forecasting System (IFS) high-resolution operational analysis (LDAS_HRES, 0.10^∘ spatial resolution) over 2017–2018. Monitoring capacities are studied by comparing open-loop and analysis experiments, again against the assimilated observations. Forecasting abilities are assessed by initializing 4 and 8 d LDAS_HRES forecasts of the LSVs with the LDAS_HRES assimilation run compared to the open-loop experiment. The positive impact of initialization from an analysis in forecast mode is particularly visible for LAI that evolves at a slower pace than SSM and is more sensitive to initial conditions than to atmospheric forcing, even at an 8 d lead time. This highlights the impact of initial conditions on LSV forecasts and the value of jointly analysing soil moisture and vegetation states.

Accurate representation of the turbulent exchange of carbon, water, and heat between the land surface and the atmosphere is critical for modelling global energy, water, and carbon cycles in both future climate projections and weather forecasts. Evaluation of models' ability to do this is performed in a wide range of simulation environments, often without explicit consideration of the degree of observational constraint or uncertainty and typically without quantification of benchmark performance expectations. We describe a Model Intercomparison Project (MIP) that attempts to resolve these shortcomings, comparing the surface turbulent heat flux predictions of around 20 different land models provided with in situ meteorological forcing evaluated with measured surface fluxes using quality-controlled data from 170 eddy-covariance-based flux tower sites. Predictions from seven out-of-sample empirical models are used to quantify the information available to land models in their forcing data and so the potential for land model performance improvement. Sites with unusual behaviour, complicated processes, poor data quality, or uncommon flux magnitude are more difficult to predict for both mechanistic and empirical models, providing a means of fairer assessment of land model performance. When examining observational uncertainty, model performance does not appear to improve in low-turbulence periods or with energy-balance-corrected flux tower data, and indeed some results raise questions about whether the energy balance correction process itself is appropriate. In all cases the results are broadly consistent, with simple out-of-sample empirical models, including linear regression, comfortably outperforming mechanistic land models. In all but two cases, latent heat flux and net ecosystem exchange of CO₂ are better predicted by land models than sensible heat flux, despite it seeming to have fewer physical controlling processes. Land models that are implemented in Earth system models also appear to perform notably better than stand-alone ecosystem (including demographic) models, at least in terms of the fluxes examined here. The approach we outline enables isolation of the locations and conditions under which model developers can know that a land model can improve, allowing information pathways and discrete parameterisations in models to be identified and targeted for future model development.

This paper introduces an ensemble square root filter (EnSRF) in the context of jointly assimilating observations of surface soil moisture (SSM) and the leaf area index (LAI) in the Land Data Assimilation System LDAS-Monde. By ingesting those satellite-derived products, LDAS-Monde constrains the Interaction between Soil, Biosphere and Atmosphere (ISBA) land surface model (LSM), coupled with the CNRM (Centre National de Recherches Météorologiques) version of the Total Runoff Integrating Pathways (CTRIP) model to improve the reanalysis of land surface variables (LSVs). To evaluate its ability to produce improved LSVs reanalyses, the EnSRF is compared with the simplified extended Kalman filter (SEKF), which has been well studied within the LDAS-Monde framework. The comparison is carried out over the Euro-Mediterranean region at a 0.25^∘ spatial resolution between 2008 and 2017. Both data assimilation approaches provide a positive impact on SSM and LAI estimates with respect to the model alone, putting them closer to assimilated observations. The SEKF and the EnSRF have a similar behaviour for LAI showing performance levels that are influenced by the vegetation type. For SSM, EnSRF estimates tend to be closer to observations than SEKF values. The comparison between the two data assimilation approaches is also carried out on unobserved soil moisture in the other layers of soil. Unobserved control variables are updated in the EnSRF through covariances and correlations sampled from the ensemble linking them to observed control variables. In our context, a strong correlation between SSM and soil moisture in deeper soil layers is found, as expected, showing seasonal patterns that vary geographically. Moderate correlation and anti-correlations are also noticed between LAI and soil moisture, varying in space and time. Their absolute value, reaching their maximum in summer and their minimum in winter, tends to be larger for soil moisture in root-zone areas, showing that assimilating LAI can have an influence on soil moisture. Finally an independent evaluation of both assimilation approaches is conducted using satellite estimates of evapotranspiration (ET) and gross primary production (GPP) as well as measures of river discharges from gauging stations. The EnSRF shows a systematic albeit moderate improvement of root mean square differences (RMSDs) and correlations for ET and GPP products, but its main improvement is observed on river discharges with a high positive impact on Nash–Sutcliffe efficiency scores. Compared to the EnSRF, the SEKF displays a more contrasting performance.

A number of studies have shown that assimilation of satellite derived soil moisture using the ensemble Kalman Filter (EnKF) can improve soil moisture estimates, particularly for the surface zone. However, the EnKF is computationally expensive since an ensemble of model integrations have to be propagated forward in time. Here, assimilating satellite soil moisture data from the Soil Moisture Active Passive (SMAP) mission, we compare the EnKF with the computationally cheaper ensemble Optimal Interpolation (EnOI) method over the contiguous United States (CONUS). The background error–covariance in the EnOI is sampled in two ways: (i) by using the stochastic spread from an ensemble open-loop run, and (ii) sampling from the model spinup climatology. Our results indicate that the EnKF is only marginally superior to one version of the EnOI. Furthermore, the assimilation of SMAP data using the EnKF and EnOI is found to improve the surface zone correlation with in situ observations at a $95\%$ significance level. The EnKF assimilation of SMAP data is also found to improve root-zone correlation with independent in situ data at the same significance level; however this improvement is dependent on which in situ network we are validating against. We evaluate how the quality of the atmospheric forcing affects the analysis results by prescribing the land surface data assimilation system with either observation corrected or model derived precipitation. Surface zone correlation skill increases for the analysis using both the corrected and model derived precipitation, but only the latter shows an improvement at the $95\%$ significance level. The study also suggests that assimilation of satellite derived surface soil moisture using the EnOI can correct random errors in the atmospheric forcing and give an analysed surface soil moisture close to that of an open-loop run using observation derived precipitation. Importantly, this shows that estimates of soil moisture could be improved using a combination of assimilating SMAP using the computationally cheap EnOI while using model derived precipitation as forcing. Finally, we assimilate three different Level-2 satellite derived soil moisture products from the European Space Agency Climate Change Initiative (ESA CCI), SMAP and SMOS (Soil Moisture and Ocean Salinity) using the EnOI, and then compare the relative performance of the three resulting analyses against in situ soil moisture observations. In this comparison, we find that all three analyses offer improvements over an open-loop run when comparing to in situ observations. The assimilation of SMAP data is found to perform marginally better than the assimilation of SMOS data, while assimilation of the ESA CCI data shows the smallest improvement of the three analysis products.

Weather centres use a variety of data assimilation schemes to analyze different land variables in their operational forecast systems. Current activities at the European Centre for Medium‐Range Weather Forecasts (ECMWF) are working towards a unified and more consistent land data assimilation system to provide more accurate initial conditions for the atmospheric forecasts. The first step is to replace the current 1D optimal interpolation (1D‐OI) used for first‐layer soil and snow temperature analyses, and integrate multi‐layer soil and first‐layer snow temperature into the ensemble‐based simplified extended Kalman filter (SEKF) currently used for multi‐layer soil moisture. This work focuses on the technical developments and the evaluation of the atmospheric forecast skill of a series of numerical weather prediction experiments to compare different SEKF configurations with the former 1D‐OI over a three‐month summer and winter period. Using the SEKF leads to seasonally varying significant improvements in the 2‐m temperature forecast in the verification against own analyses and to slightly improved results in the validation using independent synoptic observations. This work lays the foundation for integrating additional land variables into the SEKF and investigating stronger land–atmosphere coupling.