SMOS Data Assimilation for Numerical Weather Prediction

Assessment of SMOS soil moisture retrieval parameters using tau–omega algorithms for soil moisture deficit estimation

As part of the Soil Moisture and Ocean Salinity (SMOS) validation process, a comparison of the skills of three satellites [SMOS, Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) or Advanced Microwave Scanning Radiometer, and Advanced Scatterometer (ASCAT)], and one-model European Centre for Medium Range Weather Forecasting (ECMWF) soil moisture products is conducted over four watersheds located in the U.S. The four products compared in for 2010 over four soil moisture networks were used for the calibration of AMSR-E. The results indicate that SMOS retrievals are closest to the ground measurements with a low average root mean square error of 0.061 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ·m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> for the morning overpass and 0.067 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ·m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> for the afternoon overpass, which represents an improvement by a factor of 2-3 compared with the other products. The ECMWF product has good correlation coefficients (around 0.78) but has a constant bias of 0.1-0.2 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ·m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> over the four networks. The land parameter retrieval model AMSR-E product gives reasonable results in terms of correlation (around 0.73) but has a variable seasonal bias over the year. The ASCAT soil moisture index is found to be very noisy and unstable.

Long and consistent soil moisture time series at adequate spatial resolution are key to foster the application of soil moisture observations and remotely-sensed products in climate and numerical weather prediction models. The two L-band soil moisture satellite missions SMAP (Soil Moisture Active Passive) and SMOS (Soil Moisture and Ocean Salinity) are able to provide soil moisture estimates on global scales and in kilometer accuracy. However, the SMOS data record has an appropriate length of 7.5 years since late 2009, but with a coarse resolution of ∼25 km only. In contrast, a spatially-enhanced SMAP product is available at a higher resolution of 9 km, but for a shorter time period (since March 2015 only). Being the fundamental observable from passive microwave sensors, reliable brightness temperatures (Tbs) are a mandatory precondition for satellite-based soil moisture products. We therefore develop, evaluate and apply a copula-based data fusion approach for combining SMAP Enhanced (SMAP_E) and SMOS brightness Temperature (Tb) data. The approach exploits both linear and non-linear dependencies between the two satellite-based Tb products and allows one to generate conditional SMAP_E-like random samples during the pre-SMAP period. Our resulting global Copula-combined SMOS-SMAP_E (CoSMOP) Tbs are statistically consistent with SMAP_E brightness temperatures, have a spatial resolution of 9 km and cover the period from 2010 to 2018.&#13;\n&#13;\nCurrently, the data contains only horizontally polarised retrievals from SMOS's and SMAP's ascending and descending nodes, respectively.

&amp;lt;p&amp;gt;Merging data from different instruments is required to construct long time data records of soil moisture (SM). This is the goal of projects such as the ESA Climate Change Initiative (CCI) for SM (Gruber et al., 2019), which uses both active and passive microwave sensors. Currently, the GLDAS v2.1 model is used as reference to re-scale active and passive time series by matching their Cumulative Density Function (CDF) to that of the model. Removing the dependency on models is important, in particular for data assimilation applications into hydrological or climate models, and it has been proposed (Van der Schalie et al., 2018) to use L-band data from one of the two instruments specifically designed to measure SM, ESA Soil Moisture and Ocean Salinity (SMOS) and NASA Soil Moisture Active Passive (SMAP) satellites, as reference to re-scale other time series.&amp;lt;br&amp;gt;To investigate this approach, AMSR-2 SM time series obtained from C1-, C2- and X-band observations using LPRM (Land Parameter Retrieval Model) were re-scaled by CDF-matching (Brocca et al., 2011) using different SMAP and SMOS official (SMAP L2 V005, SMOS L3 V300, SMOS NRT V100&amp;amp;V200) and research (SMOS IC V103) SM products as well as the SMAP and SMOS LPRM v6 SM data used by the ESA CCI. The time series re-scaled using L-band remote sensing data were compared to those re-scaled using GLDAS and were evaluated against in situ measurements at several hundred sites retrieved from the International Soil Moisture Network (Dorigo et al., 2011). The results were analyzed as a function of the land cover class and the Koppen-Geiger climate classification.&amp;lt;br&amp;gt;Overall, AMSR-2 time series re-scaled using SMAP L2, SMAP LPRM and SMOS IC data sets as reference gave the best correlations with respect to in situ measurements, similar to those obtained by the time series re-scaled using GLDAS and slightly better than those of the original AMSR-2 time series. These results imply that different SMAP and SMOS products could actually be used to replace GLDAS as reference for the re-scaling of other sensors time series within the ESA CCI. However, one must bear in mind that this study is limited to the re-scaling of AMSR-2 data at a few hundred sites.&amp;lt;br&amp;gt;For a more detailed assessment of the L-band data set to be used for a global re-scaling, it is necessary to investigate other effects such as the spatial coverage or the time series length. SMAP spatial coverage is better than that of SMOS in regions affected by radio frequency interference. In contrast, the length of SMAP time series can be too short to capture the long term SM variability for climate applications in some regions. The CDF of SMOS time series computed from the date of SMAP launch is significantly different to those of the full length SMOS time series in some regions of the Globe. Possible ways of using a coherent SMAP/SMOS L-band data set will be discussed.&amp;lt;/p&amp;gt;

Performance evaluation of WRF-Noah Land surface model estimated soil moisture for hydrological application: Synergistic evaluation using SMOS retrieved soil moisture

In this study, we evaluated a new SMOS (Soil Moisture and Ocean Salinity) soil moisture (SM) product, developed by the collaboration of INRA (Institut National de la Recherche Agronomique) and CESBIO (Centre d'Etudes Spatiales de la BIOsphere), against the operational SMOS level 3 SM product (SMOSL3). This new product (hereinafter referred to as SMOS-INRA-CESBIO, i.e. SMOSIC in short) differs from SMOSL3 three ways: (i) the SMOSIC algorithm considers the pixel as homogeneous and does not take into account the heterogeneity of the pixel; (ii) uses a new calibration of the effective scattering albedo and soil roughness parameters; (iii) no time correlation is applied on the optical depth. The evaluation was done over North America using the (European Center for Medium range Weather Forecasting) ECMWF SM simulation as a reference, using data for 2011. A better performance of the SMOSIC SM product with respect to ECMWF was found: (i) SMOSIC had higher correlation coefficients (temporal dynamics) and lower unbiased RMSD (absolute values) values with ECMWF over most of the study area and (ii) the spatial patterns of the SMOSIC temporal mean SM maps were in a better agreement with ECMWF.

Soil moisture observations are expected to play an important role in monitoring global climate trends. However, measuring soil moisture is challenging because of its high spatial and temporal variability. Point-scale in-situ measurements are scarce and, excluding model-based estimates, remote sensing remains the only practical way to observe soil moisture at a global scale. The ESA-led Soil Moisture and Ocean Salinity (SMOS) mission, launched in 2009, measures the Earth’s surface natural emissivity at L-band and provides highly accurate soil moisture information with a 3-day revisiting time. Using the first six full annual cycles of SMOS measurements (June 2010–June 2016), this study investigates the temporal variability of global surface soil moisture. The soil moisture time series are decomposed into a linear trend, interannual, seasonal, and high-frequency residual (i.e., subseasonal) components. The relative distribution of soil moisture variance among its temporal components is first illustrated at selected target sites representative of terrestrial biomes with distinct vegetation type and seasonality. A comparison with GLDAS-Noah and ERA5 modeled soil moisture at these sites shows general agreement in terms of temporal phase except in areas with limited temporal coverage in winter season due to snow. A comparison with ground-based estimates at one of the sites shows good agreement of both temporal phase and absolute magnitude. A global assessment of the dominant features and spatial distribution of soil moisture variability is then provided. Results show that, despite still being a relatively short data set, SMOS data provides coherent and reliable variability patterns at both seasonal and interannual scales. Subseasonal components are characterized as white noise. The observed linear trends, based upon one strong El Niño event in 2016, are consistent with the known El Niño Southern Oscillation (ENSO) teleconnections. This work provides new insight into recent changes in surface soil moisture and can help further our understanding of the terrestrial branch of the water cycle and of global patterns of climate anomalies. Also, it is an important support to multi-decadal soil moisture observational data records, hydrological studies and land data assimilation projects using remotely sensed observations.

This study aims at stimulating the development of soil moisture data assimilation systems in a direction where they can provide both the necessary control of slow drift in operational NWP applications and support the physical insight in the performance of the land surface component. It addresses four topics concerning the systematic nature of soil moisture data assimilation experiments over Europe during the growing season of 2000 involving the European Centre for Medium-Range Weather Forecasts (ECMWF) model infrastructure. In the first topic the effect of the (spinup related) bias in 40-yr ECMWF Re-Analysis (ERA-40) precipitation on the data assimilation is analyzed. From results averaged over 36 European locations, it appears that about half of the soil moisture increments in the 2000 growing season are attributable to the precipitation bias. A second topic considers a new soil moisture data assimilation system, demonstrated in a coupled single-column model (SCM) setup, where precipitation and radiation are derived from observations instead of from atmospheric model fields. For many of the considered locations in this new system, the accumulated soil moisture increments still exceed the interannual variability estimated from a multiyear offline land surface model run. A third topic examines the soil water budget in response to these systematic increments. For a number of Mediterranean locations the increments successfully increase the surface evaporation, as is expected from the fact that atmospheric moisture deficit information is the key driver of soil moisture adjustment. In many other locations, however, evaporation is constrained by the experimental SCM setup and is hardly affected by the data assimilation. Instead, a major portion of the increments eventually leave the soil as runoff. In the fourth topic observed evaporation is used to evaluate the impact of the data assimilation on the forecast quality. In most cases, the difference between the control and data assimilation runs is considerably smaller than the (positive) difference between any of the simulations and the observations.

Measurements of both the Soil Moisture and Ocean Salinity (SMOS) and Aquarius L-band microwave radiometers show a significant presence of radio frequency interference (RFI), although they operate in a protected frequency band where transmission is prohibited. RFI detection and mitigation remain a challenging problem for both missions, especially for low or moderate (i.e., on the order of 10 K or less) amplitude contributions. An algorithm for low-level source detection and mitigation is already included in Aquarius data sets, and both Aquarius and SMOS have distinct attributes that can potentially enable further improvements in detection and mitigation of these sources to some degree. The combination of SMOS and Aquarius data sets may enable further future improvements as well. Initial efforts toward this goal are reported in this paper. Similarities and differences in RFI effects on SMOS and Aquarius are examined, with a particular focus on instrument properties that cause differences in received RFI power in SMOS and Aquarius observations of a specific source. A study is also performed of SMOS observations for regions reported by Aquarius to contain “low-level” RFI. It is shown that the detection of these sources in the SMOS data set is challenging and that the dependence of the SMOS third and fourth Stokes parameters on incidence angle makes the polarimetric features of SMOS difficult to utilize for low-level source detection. However, an angular fitting procedure suggested previously in the literature can, in some cases, detect such sources in horizontal and vertical polarizations.

ESA's Soil Moisture and Ocean Salinity (SMOS) mission has been designed to extend our knowledge of the Earth's water cycle. Soil Moisture and Ocean Salinity records brightness temperatures at the L‐band, which over land are sensitive to soil and vegetation parameters. On the basis of these measurements, soil moisture and vegetation opacity data sets have been derived operationally since 2009 for applications comprising hydrology, numerical weather prediction (NWP), and drought monitoring. We present a method to enhance the knowledge about the temporal evolution of radiative transfer parameters. The radiative transfer model L‐Band Microwave Emission of the Biosphere (L‐MEB) is used within a data assimilation framework to retrieve vegetation opacity and soil surface roughness. To analyze the ability of the data assimilation approach to track the temporal evolution of these parameters, scenario analyses were performed with increasing complexity. First, the HYDRUS‐1D code was used to generate soil moisture and soil temperature time series. On the basis of these data, the L‐MEB forward model was run to simulate brightness temperature observations. Finally, the coupled model system HYDRUS‐1D and L‐MEB were integrated into a data assimilation framework using a particle filter, which is able to update L‐MEB states as well as L‐MEB parameters. Time invariant and time variable radiative transfer parameters were estimated. Moreover, it was possible to estimate a “bias” term when model simulations show a systematic difference as compared to observations. An application to a USDA‐NRCS Soil Climate Analysis Network (SCAN) site showed the good performance of the proposed approach under real conditions.

The Soil Moisture and Ocean Salinity (SMOS) satellite is a 2D interferometer in L-band (1.4 GHz). Over land, it enables to probe the earth surface emissivity related to soil moisture in the first centimeters of the soil, with an aimed accuracy better than 4m³.m^-3 and an average spatial resolution of 40km. The European Space Agency's (ESA) ground segment provides half-orbit soil moisture products at level 2. The Centre National d'Etudes Spatiales (CNES) in France has developed the CATDS (Centre Aval de Traitement des Données SMOS) ground segment in order to produce global maps, known as level 3 and 4 products. Over land, the algorithm is based on the level 2 soil moisture ESA's prototype. The major enhancement of the CATDS concerns the use of multi-orbit retrieval. The level 3 Soil moisture (SM) products are global maps of soil moisture, and other geophysical products (vegetation optical thickness, albedo, soil dielectric constant or surface temperature). For a particular point, the revisit time is between 1 and 3 days, and the entire Earth's surface is covered by SMOS field of view in 3 days. Level 3 SM products are available over different time periods. First, a 1 day global product is generated for each day. Then, 1 day global maps are aggregated in 3-day global products, 10- day and monthly products.

The Soil Moisture and Ocean Salinity (SMOS) mission delivers global surface soil moisture fields at high temporal resolution which is of major relevance for water management and climate predictions. Between April 26 and May 9, 2010, an airborne campaign with the L-band radiometer EMIRAD-2 was carried out within one SMOS pixel (44 × 44 km) in the Skjern River Catchment, Denmark. Concurrently, ground sampling was conducted within three 2 × 2 km patches (EMIRAD footprint size) of differing land cover. By means of this data set, the objective of this study is to present the validation of SMOS L1C brightness temperatures <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TB</i> of the selected node. Data is stepwise compared from point via EMIRAD to SMOS scale. From ground soil moisture samples, <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TB</i> 's are pointwise estimated through the L-band microwave emission of the biosphere model using land cover specific model settings. These <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TB</i> 's are patchwise averaged and compared with EMIRAD <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TB</i> 's. A simple uncertainty assessment by means of a set of model runs with the most influencing parameters varied within a most likely interval results in a considerable spread of <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TB</i> 's (5-20 K). However, for each land cover class, a combination of parameters could be selected to bring modeled and EMIRAD data in good agreement. Thereby, replacing the Dobson dielectric mixing model with the Mironov model decreases the overall RMSE from 11.5 K to 3.8 K. Similarly, EMIRAD data averaged at SMOS scale and corresponding SMOS <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TB</i> 's show good accordance on the single day where comparison is not prevented by strong radio-frequency interference (RFI) (May 2, avg. RMSE = 9.7 K). While the advantages of solid data sets of high spatial coverage and density throughout spatial scales for SMOS validation could be clearly demonstrated, small temporal variability in soil moisture conditions and RFI contamination throughout the campaign limited the extent of the validation work. Further attempts over longer time frames are planned by means of soil moisture network data as well as studies on the impacts of organic layers under natural vegetation and higher open water fractions at surrounding grid nodes.

Soil moisture is an important hydrologic variable due to its influence on plant growth and the surface energy budget. Most current soil moisture networks measure soil moisture at single points; these point measurements of soil moisture may not relate to the soil moisture of the larger, surrounding areas. Crop models and weather and climate models, which are sensitive to soil moisture, could be improved by having accurate soil moisture data that is relevant at scales larger than a point. Microwave C–band and X–band sensing satellites have attempted to retrieve soil moisture at relevant, large spatial scales needed for models, but these satellites are limited in their sensing depth, sensitivity to vegetation, and sensitivity to scattering by precipitation. Two recent satellite missions, the European Space Agency’s Soil Moisture and Ocean Salinity (SMOS) mission and NASA’s Soil Moisture Active Passive (SMAP) mission, retrieve soil moisture from L–band measurements of radiation, which improve upon the limitations of previous satellites. The retrieval of soil moisture from microwave satellite measurements requires knowledge of the vegetation optical thickness (τ), which is related to the vegetation water content (VWC). The multi-incidence angular design of SMOS provides its retrieval algorithm the advantage of retrieving soil moisture and τ simultaneously. SMAP will measure at a single incidence angle and will require an ancillary source of τ data. The current plan is for SMAP to use a climatology of normalized difference vegetation index (NDVI) data to generate a τ climatology. This dissertation analyzed this SMAP τ climatology by first validating SMOS τ and then by comparing SMOS τ to SMAP τ . The validation of SMOS τ was done by comparing SMOS τ during the growing season to crop yield estimates for each Iowa county. The result was that larger values of SMOS τ were correlated with larger crop yields, which suggested that SMOS τ is related to the growth of vegetation. SMOS τ was also found to be noisy and to behave strangely outside of the growing

Estimation of soil moisture at large scale has been performed using several satellite-based passive microwave sensors and a variety of retrieval methods over the past two decades. The most recent source of soil moisture is the European Space Agency Soil Moisture and Ocean Salinity (SMOS) mission. A thorough validation must be conducted to insure product quality that will, in turn, support the widespread utilization of the data. This is especially important since SMOS utilizes a new sensor technology and is the first passive L-band system in routine operation. In this paper, we contribute to the validation of SMOS using a set of four in situ soil moisture networks located in the U.S. These ground-based observations are combined with retrievals based on another satellite sensor, the Advanced Microwave Scanning Radiometer (AMSR-E). The watershed sites are highly reliable and address scaling with replicate sampling. Results of the validation analysis indicate that the SMOS soil moisture estimates are approaching the level of performance anticipated, based on comparisons with the in situ data and AMSR-E retrievals. The overall root-mean-square error of the SMOS soil moisture estimates is 0.043 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> for the watershed networks (ascending). There are bias issues at some sites that need to be addressed, as well as some outlier responses. Additional statistical metrics were also considered. Analyses indicated that active or recent rainfall can contribute to interpretation problems when assessing algorithm performance, which is related to the contributing depth of the satellite sensor. Using a precipitation flag can improve the performance. An investigation of the vegetation optical depth (tau) retrievals provided by the SMOS algorithm indicated that, for the watershed sites, these are not a reliable source of information about the vegetation canopy. The SMOS algorithms will continue to be refined as feedback from validation is evaluated, and it is expected that the SMOS estimates will improve.

In this study, we compared coupled and uncoupled land surface model simulations of surface soil moisture (SM) with passive microwave remote sensing observations of SM over the contiguous US (CONUS). For this purpose, we used the following: (i) the European Space Agency's Soil Moisture and Ocean Salinity (SMOS) satellite retrieved L3 surface SM; (ii) the Advanced Microwave Scanning Radiometer for EOS (AMSR-E) SM derived using LPRM; and (iii) the coupled and uncoupled ORCHIDEE land surface model. The comparison was achieved by computing the temporal mean difference between the normalized remotely-sensed and the model-SM datasets. It was found that both coupled and uncoupled models are drier than the remotely sensed data (particularly the SMOS data) over the principal aquifers of the CONUS. These aquifers are known as one of the places where irrigation is intensive. Therefore, the reason behind this model’ dryness could be the fact that the models do not take into account the irrigation activities, which can be monitored by the remotely-sensed observations. Time series of SM over irrigated pixels also showed that SMOS was wetter than the models mainly during the irrigation period.