Soil Moisture Retrieval Algorithm Research Articles

Improving the AMSR-E/NASA Soil Moisture Data Product Using In-Situ Measurements from the Tibetan Plateau

The daily AMSR-E/NASA (the Advanced Microwave Scanning Radiometer-Earth Observing System/the National Aeronautics and Space Administration) and JAXA (the Japan Aerospace Exploration Agency) soil moisture (SM) products from 2002 to 2011 at 25 km resolution were developed and distributed by the NASA National Snow and Ice Data Center Distributed Active Archive Center (NSIDC DAAC) and JAXA archives, respectively. This study analyzed and evaluated the temporal changes and accuracy of the AMSR-E/NASA SM product and compared it with the AMSR-E/JAXA SM product. The accuracy of both AMSR-E/NASA and JAXA SM was low, with RMSE (root mean square error) > 0.1 cm3 cm−3 against the in-situ SM measurements, especially the AMSR-E/NASA SM. Compared with the AMSR-E/JAXA SM, the dynamic range of AMSR-E/NASA SM is very narrow in many regions and does not reflect the intra- and inter-annual variability of soil moisture. We evaluated both data products by building a linear relationship between the SM and the Microwave Polarization Difference Index (MPDI) to simplify the AMSR-E/NASA SM retrieval algorithm on the basis of the observed relationship between samples extracted from the MPDI and SM data. We obtained the coefficients of this linear relationship (i.e., A0 and A1) using in-situ measurements of SM and brightness temperature (TB) data simulated with the same radiative transfer model applied to develop the AMSR-E/NASA SM algorithm. Finally, the linear relationships between the SM and MPDI were used to retrieve the SM monthly from AMSR-E TB data, and the estimated SM was validated using the in-situ SM measurements in the Naqu area on the Tibetan Plateau of China. We obtained a steeper slope, i.e., A1 = 8, with the in-situ SM measurements against A1 = 1, when using the NASA SM retrievals. The low A1 value is a measure of the low sensitivity of the NASA SM retrievals to MPDI and its narrow dynamic range. These results were confirmed by analyzing a data set collected in Poland. In the case of the Tibetan Plateau, the higher value A1 = 8 gave more accurate monthly AMSR-E SM retrievals with RMSE = 0.065 cm3 cm−3. The dynamic range of the improved retrievals was more consistent with the in-situ SM measurements than with both the AMSR-E/NASA and JAXA SM products in the Naqu area of the Tibetan Plateau in 2011.

Impact of temporal variations in vegetation optical depth and vegetation temperature on L-band passive soil moisture retrievals over a tropical forest using in-situ information

ABSTRACTThe Soil Moisture and Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) missions provide estimates of soil moisture (SM) at similar spatial resolutions using L-band brightness temperatures (). These missions meet the requirement of SM retrievals with an unbiased root-mean-square difference (ubRMSD) 0.04 (m m−3) compared to in-situ measurements over most of the ecosystems; however, their SM estimates over forests present an ubRMSD 0.10 (m m−3). In this paper, we compared the SM retrievals from the SMOS and SMAP SM products with in-situ SM over a tropical forest in Southern Mexico. The L-band passive SM retrievals were evaluated in terms of four statistical metrics: root-mean-square difference (RMSD), bias, ubRMSD, and correlation coefficient (). In-situ measurements of SM, soil and vegetation temperatures, precipitation, soil surface roughness, tree heights, diameters at the breast height of trunks, and forest cover fraction were collected during a field campaign from 6 January to 14 June 2015 in the Biosphere Reserve of Calakmul, Mexico, covering two areas of about 40 km 40 km each. The comparison between SM retrievals from SMOS and SMAP and in-situ SM showed an RMSD ranging from 0.107 to 0.322 (m m−3) and an ubRMSD from 0.049 to 0.128 (m m−3). Overall, the SMAP SM estimates showed higher values of and were closer to in-situ SM. Because the SMAP and SMOS radiometers performed very similar, these differences are due to the values assigned to the vegetation optical depth (), the scattering albedo (), and the representation of the dynamics in vegetation and soil temperatures in the SMOS and SMAP retrieval algorithm. Based on an optimization process, we estimated simultaneously optimal and values for the tropical forest by using observations from SMAP and SMOS radiometers.The optimal value of was 0.0655 for the tropical forest, and constant over the study period. In contrast, the optimal values of showed to be variant on time and ranging between 1.0 and 1.7, with an averaged value of 1.4 and a standard deviation of 0.24. When applying the optimal values of and and in-situ soil and vegetation temperatures, the SM retrievals showed an ubRMSD of 0.035–0.070 (m m), improving the SM retrievals about 45%. A sensitivity analysis was conducted to evaluate the effect of the uncertainties in , , and soil and vegetation temperatures on the estimates of . It was found that vegetation temperature () was the most sensitive parameter, with higher than 0.70 for both polarizations in . When comparing in-situ and surface temperature values used in the SMAP and SMOS SM retrieval algorithms, differences up to 10 K were observed, affecting the SM estimates. The results presented in this paper could be useful in the preparation of the SMAP Calibration/Validation Experiment 2019, aiming at improving SM retrievals over forests.

High Spatio-Temporal Resolution CYGNSS Soil Moisture Estimates Using Artificial Neural Networks

This paper presents a learning-based, physics-aware soil moisture (SM) retrieval algorithm for NASA’s Cyclone Global Navigation Satellite System (CYGNSS) mission. The goal of the proposed novel method is to advance CYGNSS-based SM estimations, exploiting the spatio-temporal resolution of the GNSS reflectometry (GNSS-R) signals to its highest potential within a machine learning framework. The methodology employs a fully connected Artificial Neural Network (ANN) regression model to perform SM predictions through learning the nonlinear relations of SM and other land geophysical parameters to the CYGNSS observables. In situ SM measurements from several International SM Network (ISMN) sites are used as reference labels; CYGNSS incidence angles, derived reflectivity and trailing edge slope (TES) values, as well as ancillary data, are exploited as input features for training and validation of the ANN model. In particular, the utilized ancillary data consist of normalized difference vegetation index (NDVI), vegetation water content (VWC), terrain elevation, terrain slope, and h-parameter (surface roughness). Land cover classification and inland water body masks are also used for the intermediate derivations and quality control purposes. The proposed algorithm assumes uniform SM over a 0.0833 ∘ × 0.0833 ∘ (approximately 9 km × 9 km around the equator) lat/lon grid for any CYGNSS observation that falls within this window. The proposed technique is capable of generating sub-daily and high-resolution SM predictions as it does not rely on time-series or spatial averaging of the CYGNSS observations. Once trained on the data from ISMN sites, the model is independent from other SM sources for retrieval. The estimation results obtained over unseen test data are promising: SM predictions with an unbiased root mean squared error of 0.0544 cm 3 /cm 3 and Pearson correlation coefficient of 0.9009 are reported for 2017 and 2018.

Spatio-Temporal Mapping of L-Band Microwave Emission on a Heterogeneous Area with ELBARA III Passive Radiometer

Water resources on Earth become one of the main concerns for society. Therefore, remote sensing methods are still under development in order to improve the picture of the global water cycle. In this context, the microwave bands are the most suitable to study land–water resources. The Soil Moisture and Ocean Salinity (SMOS), satellite mission of the European Space Agency (ESA), is dedicated for studies of the water in soil over land and salinity of oceans. The part of calibration/validation activities in order to improve soil moisture retrieval algorithms over land is done with ground-based passive radiometers. The European Space Agency L-band Microwave Radiometer (ELBARA III) located near the Bubnów wetland in Poland is capable of mapping microwave emissivity at the local scale, due to the azimuthal and vertical movement of the horn antenna. In this paper, we present results of the spatio-temporal mapping of the brightness temperatures on the heterogeneous area of the Bubnów test-site consisting of an area with variable organic matter (OM) content and different type of vegetation. The soil moisture (SM) was retrieved with the L-band microwave emission of the biosphere (L-MEB) model with simplified roughness parametrization (SRP) coupling roughness and optical depth parameters. Estimated soil moisture values were compared with in-situ data from the automatic agrometeorological station. The results show that on the areas with a relatively low OM content (4–6%—cultivated field) there was good agreement between measured and estimated SM values. Further increase in OM content, starting from approximately 6% (meadow wetland), caused an increase in bias, root mean square error (RMSE), and unbiased RMSE (ubRMSE) values and a general drop in correlation coefficient (R). Despite a span of obtained R values, we found that time-averaged estimated SM using the L-MEB SRP approach strongly correlated with OM contents.

Closed-Form Anisotropic Polarimetric Two-Scale Model for Fast Evaluation of Sea Surface Backscattering

The polarimetric two-scale model (PTSM) was introduced a few years ago as an electromagnetic scattering model to be used within algorithms for soil moisture retrieval from polarimetric synthetic aperture radar (SAR) data. PTSM inherits the ability to account for depolarization effects from the original two-scale model (TSM), and, with respect to the latter, it has the advantage to provide closed-form expressions of the elements of the covariance matrix that hold for moderate large-scale surface slopes. This allows a very fast evaluation of scattering, since numerical integration needed by the original TSM is avoided. The TSM, also called composite model (CM), has been extensively used to study scattering from the sea surface, so that it is natural to explore the use of PTSM for the same purpose. However, in its current formulation, PTSM assumes that large-scale surface slope distribution and small-scale roughness spectrum are isotropic. This is not realistic for the sea surface, for which anisotropy is dictated by the wind direction. Accordingly, we here extend PTSM to account for surface roughness anisotropy, so obtaining the anisotropic PTSM (A-PTSM). In addition, as a second contribution, we provide A-PTSM expressions also in the circular polarization basis, which may be useful for some SAR sensor polarimetric configurations. Finally, we compare A-PTSM results with sea surface scattering measurements available in the literature and with results of the second-order small-slope approximation (SSA2). In particular, as a third original contribution of this paper, an analytical closed-form expression of the ratio of crosspolarized normalized radar cross sections (NRCSs) obtained by SSA2 and A-PTSM is provided.

Using Cosmic-Ray Neutron Probes in Validating Satellite Soil Moisture Products and Land Surface Models

Soil moisture content is one of the most important parameters of hydrological studies. Cosmic-ray neutron sensing is a promising proximal soil moisture sensing technique at intermediate scale and high temporal resolution. In this study, we validate satellite soil moisture products for the period of March 2015 and December 2018 by using several existing Cosmic Ray Neutron Probe (CRNP) stations of the COSMOS database and a CRNP station that was installed in the south part of Turkey in October 2016. Soil moisture values, which were inferred from the CRNP station in Turkey, are also validated using a time domain reflectometer (TDR) installed at the same location and soil water content values obtained from a land surface model (Noah LSM) at various depths (0.1 m, 0.3 m, 0.6 m and 1.0 m). The CRNP has a very good correlation with TDR where both measurements show consistent changes in soil moisture due to storm events. Satellite soil moisture products obtained from the Soil Moisture and Ocean Salinity (SMOS), the METOP-A/B Advanced Scatterometer (ASCAT), Soil Moisture Active Passive (SMAP), Advanced Microwave Scanning Radiometer 2 (AMSR2), Climate Change Initiative (CCI) and a global land surface model Global Land Data Assimilation System (GLDAS) are compared with the soil moisture values obtained from CRNP stations. Coefficient of determination ( r 2 ) and unbiased root mean square error (ubRMSE) are used as the statistical measures. Triple Collocation (TC) was also performed by considering soil moisture values obtained from different soil moisture products and the CRNPs. The validation results are mainly influenced by the location of the sensor and the soil moisture retrieval algorithm of satellite products. The SMAP surface product produces the highest correlations and lowest errors especially in semi-arid areas whereas the ASCAT product provides better results in vegetated areas. Both global and local land surface models’ outputs are highly compatible with the CRNP soil moisture values.

Calibration of SMOS Soil Moisture Retrieval Algorithm: A Case of Tropical Site in Malaysia

Soil Moisture and Ocean Salinity (SMOS) mission has successfully contributed to global soil moisture products since 2009. Validation and calibration activities were conducted worldwide, yet some of the validation results do not fulfill the targeted accuracy of ±0.04 $\text{m}^{3}\text{m}^{-3}$ . This paper presented the site-specific calibration of the V620 retrieval algorithm with in situ data collected at selected agricultural sites in the humid tropical regions, Malaysia. This set of data has been validated where low accuracy of SMOS soil moisture products was found. To improve the SMOS soil moisture retrieval, calibration of SMOS soil moisture retrieval algorithm based on the L-band Microwave Emission and Biosphere model and SMOS Level 1C $\text{T}_{\mathrm {B}}$ products, considering the local parameters was conducted. The calibration proves that these site-specific parameters improve the product’s accuracy. Validation of SMOS Level 2 product with in situ data showed bias, root-mean-square error (RMSE), and unbiased RMSE (ubRMSE) ranging from 0.050 to 0.118 $\text{m}^{3}\text{m}^{-3}$ , 0.068 to 0.142 $\text{m}^{3}\text{m}^{-3}$ , and 0.069 to 0.103 $\text{m}^{3}\text{m}^{-3}$ , respectively. The soil moisture retrieval based on the calibrated model showed an improved bias of 0.020–0.056 $\text{m}^{3}\text {m}^{-3}$ and RMSE of 0.026–0.065 $\text{m}^{3}\text{m}^{-3}$ . The ubRMSE ranges from 0.017 to 0.034 $\text{m}^{3}\text{m}^{-3}$ . Recently released SMOS-IC V105 product was also validated, where small improvements were noticed when compared to the accuracy of SMOS Level 2. This paper shows the importance of local parameters in retrieving soil moisture with higher accuracy compared to the use of global generalized parameters that are used in the original SMOS soil moisture retrieval algorithm.

Passive/active microwave soil moisture change disaggregation using SMAPVEX12 data

The SMAPVEX12 (Soil Moisture Active Passive (SMAP) Validation Experiment 2012) experiment was conducted during June-July 2012 in Manitoba, Canada with the goal of collecting remote sensing data and ground measurements for the development and testing of soil moisture retrieval algorithms under varying vegetation and soil conditions for the SMAP satellite. The aircraft based soil moisture data provided by the passive/active microwave sensor PALS (Passive and Active L-band System) has a nominal spatial resolution of 1600 m. However, this resolution is not compatible with agricultural, meteorological and hydrological studies that require high spatial resolutions and this issue can be solved by soil moisture disaggregation. The soil moisture disaggregation algorithm integrates radiometer soil moisture retrievals and high-resolution radar observations and it can provide soil moisture estimates at a finer scale than the radiometer data alone. In this study, a change detection algorithm was used for disaggregation of coarse resolution passive microwave soil moisture retrievals with radar backscatter coefficients obtained from the higher spatial resolution UAVSAR (Unmanned Air Vehicle Synthetic Aperture Radar) at crop field scale. The accuracy of the disaggregated change in soil moisture was evaluated using ground based soil moisture measurements collected during SMAPVEX12 campaign. The results showed that soil moisture spatial variabilities were better characterized by the disaggregated change in soil moisture estimates at 5 m/800 m resolution as well as a good agreement between disaggregated estimates and in situ measurements. It also showed that VWC (Vegetation Water Content) did not have a big impact on disaggregation algorithm performance, with R2 of the disaggregated results ranging 0.628–0.794. The 5 m and 800 m resolution disaggregated soil moisture did no show significant difference in statistical performance variables.

Uncertainty evaluation at three spatial scales for the NDVI-based VWC estimation method used in the SMAP algorithm

ABSTRACTVegetation water content (VWC) is the key input parameter for a soil moisture retrieval algorithm based on microwave remote sensing, and VWC uncertainty can limit the estimated accuracy of soil moisture. There has been little research on VWC algorithm development and validation in China, and the uncertainty of the VWC estimation method has not been well evaluated. Therefore, the aim of this study is to evaluate the uncertainty of the VWC estimation method used in the SMAP (Soil Moisture Active Passive) algorithm on three spatial scales (the point-scale, 30 m scale, and 1 km scale) for maize in northeast China. Results from three ground experimental datasets showed that the SMAP VWC estimation method was strongly biased with an average overestimation of 1.16 kg m−2,1.04 kg m−2, and 1.13 kg m−2 for the point-scale, 30 m scale, and 1 km scale respectively, and maximum bias occurred in the mid-stage of maize. Also, a new power relationship between NDVI (Normalized Difference Vegetation Index) and VWC was proposed for the 30 m scale based on Sentinel 2 NDVI and field VWC values from 2017 experiment, with respective R2 (coefficient of determination) and Root Mean Squared Error (RMSE) values of 0.80 and 0.67 kg m−2. The results confirmed that this power relationship was still suitable for VWC estimation at the 1 km scale, and it has smaller bias than the original SMAP VWC method. Future work will be carried out to evaluate the applicability of this VWC estimation method over a lager region. It is expected that it can improve the accuracy of soil moisture by providing high precision VWC input parameters.

Soil Moisture Retrieval Algorithm Based on TFA and CNN

With the development of mission critical sensors and sensor networks (MC-SSNs), using MC sensor for monitoring the changes of soil moisture has become convenient and real time. However, the method used to process the data obtained by the MC sensor directly affects the monitoring effect. Although the past research methods have achieved certain results, they all need to extract data features. In this paper, we propose the soil moisture retrieval algorithm with time–frequency analysis and convolutional neural network (CNN) based on ultra-wideband radar echoes, which do not need to build feature database in advance. The algorithm transforms the soil echoes into time–frequency (TF) distribution patterns and utilizes the CNN algorithm to classify soil echoes with different soil volumetric water contents (VWCs). Wigner–Ville (WV) distribution and Choi–Williams (CW) distribution are the two methods used for time–frequency transform; VGGNet and AlexNet are, respectively, applied to classify the TF patterns with different VWCs. We totally construct four soil moisture retrieval systems (WV-AlexNet, CW-AlexNet, WV-VGGNet, and CW-VGGNet), and the echoes of 27 soil water contents with different signal-to-noise ratios (SNRs) are studied. The simulation results with raw data show that the correct recognition rate of soil VWCs can reach 100% when the soil echoes are at 10 dB (SNR). The WV-AlexNet system has the best recognition performance among the four systems.

Soil Moisture in the Biebrza Wetlands Retrieved from Sentinel-1 Imagery

The objective of the study was to estimate soil moisture (SM) from Sentinel-1 (S-1) satellite images acquired over wetlands. The study was carried out during the years 2015–2017 in the Biebrza Wetlands, situated in north-eastern Poland. At the Biebrza Wetlands, two Sentinel-1 validation sites were established, covering grassland and marshland biomes, where a network of 18 stations for soil moisture measurement was deployed. The sites were funded by the European Space Agency (ESA), and the collected measurements are available through the International Soil Moisture Network (ISMN). The SAR data of the Sentinel-1 satellite with VH (vertical transmit and horizontal receive) and VV (vertical transmit and vertical receive) polarization were applied to SM retrieval for a broad range of vegetation and soil moisture conditions. The methodology is based on research into the effect of vegetation on backscatter (σ°) changes under different soil moisture and Normalized Difference Vegetation Index (NDVI) values. The NDVI was derived from the optical imagery of a MODIS (Moderate Resolution Imaging Spectroradiometer) sensor onboard the Terra satellite. It was found that the state of the vegetation expressed by NDVI can be described by the indices such as the difference between σ° VH and VV, or the ratio of σ° VV/VH, as calculated from the Sentinel-1 images in the logarithmic domain. The most significant correlation coefficient for soil moisture was found for data that was acquired from the ascending tracks of the Sentinel-1 satellite, characterized by the lowest incidence angle, and SM at a depth of 5 cm. The study demonstrated that the use of the inversion approach, which was applied to the newly developed models using Water Cloud Model (WCM) that includes the derived indices based on S-1, allowed the estimation of SM for wetlands with reasonable accuracy (10 vol. %). The developed soil moisture retrieval algorithms based on S-1 data are suited for wetland ecosystems, where soil moisture values are several times higher than in agricultural areas.

The Potential Utility of Satellite Soil Moisture Retrievals for Detecting Irrigation Patterns in China

Climate change and anthropogenic activities, including agricultural irrigation have significantly altered the global and regional hydrological cycle. However, human-induced modification to the natural environment is not well represented in land surface models (LSMs). In this study, we utilize microwave-based soil moisture products to aid the detection of under-represented irrigation processes throughout China. The satellite retrievals used in this study include passive microwave observations from the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) and its successor AMSR2, active microwave observations from the Advanced Scatterometer (ASCAT), and the blended multi-sensor soil moisture product from the European Space Agency (i.e., ESA CCI product). We first conducted validations of the three soil moisture retrievals against in-situ observations (collected from the nationwide agro-meteorological network) in irrigated areas in China. It is found that compared to the conventional Spearman’s rank correlation and Pearson correlation coefficients, entropy-based mutual information is more suitable for evaluating soil moisture anomalies induced by irrigation. In general, around 60% of uncertainties in the anomaly of “ground truth” time series can be resolved by soil moisture retrievals, with ASCAT outperforming the others. Following this, the potential utility of soil moisture retrievals in mapping irrigation patterns in China is investigated by examining the difference in probability distribution functions (detected by two-sample Kolmogorov-Smirnov test) between soil moisture retrievals and benchmarks of the numerical model ERA-Interim without considering the irrigation process. Results show that microwave remote sensing provides a promising alternative to detect the under-represented irrigation process against the reference LSM ERA-Interim. Specifically, the highest performance in detecting irrigation intensity is found when using ASCAT in Huang-Huai-Hai Plain, followed by advanced microwave scanning radiometer (AMSR) and ESA CCI. Compared to ASCAT, the irrigation detection capabilities of AMSR exhibit higher discrepancies between descending and ascending orbits, since the soil moisture retrieval algorithm of AMSR is based on surface temperature and, thus, more affected by irrigation practices. This study provides insights into detecting the irrigation extent using microwave-based soil moisture with aid of LSM simulations, which has great implications for numerical model development and agricultural managements across the country.

Soil Moisture Retrieval from the Chinese GF-3 Satellite and Optical Data over Agricultural Fields.

Timely and accurate soil moisture information is of great importance in agricultural monitoring. The Gaofen-3 (GF-3) satellite, the first C-band multi-polarization synthetic-aperture radar (SAR) satellite in China, provides valuable data sources for soil moisture monitoring. In this study, a soil moisture retrieval algorithm was developed for the GF-3 satellite based on a backscattering coefficient simulation database. We adopted eight optical vegetation indices to determine the relationships between these indices and vegetation water content (VWC) by combining Landsat-8 data and field measurements. A backscattering coefficient database was built using an advanced integral equation model (AIEM). The effects of vegetation on backscattering coefficients were corrected using the water cloud model (WCM) to obtain the bare soil backscattering coefficient (). Then, soil moisture retrievals were obtained at HH, VV and HH+VV combination respectively by minimizing the observed bare soil backscattering coefficient () and the AIEM-simulated backscattering coefficient (). Finally, the proposed algorithm was validated in agriculture region of wheat and corn in China using ground soil moisture measurements. The results showed that the normalized difference infrared index (NDII) had the best fit with measured VWC values (R = 0.885) among the eight vegetation water indices; thus, it was adopted to correct the effects of vegetation. The proposed algorithm using GF-3 satellite data performed well in soil moisture retrieval, and the scheme combining HH and VV polarization exhibited the highest accuracy, with a root mean square error (RMSE) of 0.044 m3m−3, followed by HH polarization (RMSE = 0.049 m3m−3) and VV polarization (RMSE = 0.053 m3m−3). Therefore, the proposed algorithm has good potential to operationally estimate soil moisture from the new GF-3 satellite data.

Improved surface soil moisture anomalies from Fengyun-3B over the Jiangxi province of the People’s Republic of China

ABSTRACTThis study develops a data-driven modification scheme for a commonly used soil moisture retrieval algorithm by introducing a vegetation density-related single scattering albedo based on in situ and Fengyun-3B passive microwave observations. The Jiangxi province in China’s mainland is one of the most challenging regions for soil moisture retrievals due to its complex topography, open water, and vegetation conditions. However, it has a very dense in situ soil moisture observation network which makes it a suitable test-bed to examine the performance of the modification scheme. The development of this new scheme consists of two steps. In a first step, the model is initialized using the most recently developed algorithm configuration. In the second step, these initial outcomes are used as input to determine the vegetation density related single scattering albedo which is solely based on observational data and used as the final algorithm configuration over our study area. We start by comparing the two most recent algorithm configurations against the in situ soil moisture network and demonstrate an overall improvement in terms of correlations coefficient for the most recent version. Then, the observational data- driven modification scheme was proposed and evaluated against the in situ soil moisture network with further improvements after its implementation. We furthermore applied the vegetation density-based scattering albedo in soil moisture retrievals over all grid cells in Jiangxi, and found that soil moisture data with the newly developed configuration significantly improved (up to 30%) compared to the preceding algorithm configurations. The two existing algorithm configurations were also evaluated over all grid cells and all results indicate consistent improvements between the successive algorithm versions.

Assessment of the SMAP Soil Emission Model and Soil Moisture Retrieval Algorithms for a Tibetan Desert Ecosystem

The Soil Moisture Active Passive (SMAP) satellite mission launched in January 2015 provides worldwide soil moisture (SM) monitoring based on L-band brightness temperature ( $T_{B}^{\mathrm{ p}}$ ) measurements at horizontal ( $\text{T}_{B}^{\mathrm{ H}}$ ) and vertical ( $T_{B}^{\mathrm{ V}}$ ) polarizations. This paper presents a performance assessment of SMAP soil emission model and SM retrieval algorithms for a Tibetan desert ecosystem. It is found that the SMAP emission model largely underestimates the SMAP measured $T_{B}^{\mathrm{ H}}$ (≈ 15 K), and the $T_{B}^{\mathrm{ V}}$ is underestimated during dry-down episodes. A cold bias is noted for the SMAP effective temperature due to underestimation of soil temperature, leading to the $T_{B}^{\mathrm{ p}}$ underestimation (>5 K). The remaining $T_{B}^{\mathrm{ H}}$ underestimation is found to be related to the surface roughness parameterization that underestimates its effect on modulating the $T_{B}^{\mathrm{ p}}$ measurements. Further, the topography and uncertainty of soil information are found to have minor impacts on the $T_{B}^{\mathrm{ p}}$ simulations. The SMAP baseline SM products produced by single-channel algorithm (SCA) using the $T_{B}^{\mathrm{ V}}$ measurements capture the measured SM dynamics well, while an underestimation is noted for the dry-down periods because of $T_{B}^{\mathrm{ V}}$ underestimation. The products based on the SCA with $T_{B}^{\mathrm{ H}}$ measurements underestimate the SM due to underestimation of $T_{B}^{\mathrm{ H}}$ , and the dual-channel algorithm overestimates the SM. After implementing a new surface roughness parameterization and improving the soil temperature and texture information, the deficiencies noted above in $T_{B}^{\mathrm{ p}}$ simulation and SM retrieval are greatly resolved. This indicates that the SMAP SM retrievals can be enhanced by improving both surface roughness and adopted soil temperature and texture information for Tibetan desert ecosystem.

Surface rock effects on soil moisture retrieval from L-band passive microwave observations

The L-band (1.41 GHz) passive microwave remote sensing technique is the approach used by the first satellites dedicated to soil moisture measurement, the European Space Agency's (ESA) Soil Moisture and Ocean Salinity (SMOS), and the Soil Moisture Active Passive (SMAP) mission developed by the National Aeronautics and Space Administration (NASA). These satellites aim to provide global soil moisture maps for the top ~5 cm layer of soil with an accuracy better than 0.04 m3/m3. However, with a passive microwave observing resolution of ~40 km, non-soil targets such as surface rock may possibly confound the brightness temperature observations and degrade the accuracy of retrievals for many SMOS and SMAP pixels across the world. Since the microwave contribution of rock is not well accounted for in current soil moisture retrieval algorithms, simply ignoring its existence may be detrimental to the performance of resultant soil moisture products. Using a combination of model simulations and airborne field campaign data from central Australia, this study has determined that a rock cover fraction threshold of up to 0.4 can be tolerated before the 0.04 m3/m3 soil moisture target accuracy is potentially exceeded under extreme dry or wet conditions. However, this threshold reduces to 0.2 when assessed in terms of a brightness temperature impact >4 K. These rock fraction thresholds have subsequently been applied to the Ecoclimap rock cover map, identifying the SMOS and SMAP pixels globally that are likely to be adversely affected if rock is unaccounted for. The results show that approximately ~3.3% of all SMOS and SMAP pixels may have brightness temperature impacts exceeding 4 K from surface rock, with Asia being the most affected, having ~6.0% affected pixels. These values reduce to ~1.5% of SMOS and SMAP pixels globally, and ~3.1% for Asia, when assessed in terms of soil moisture errors expected to possibly exceed 0.04 m3/m3 when not accounting for surface rock.

Impact of surface roughness, vegetation opacity and soil permittivity on L-band microwave emission and soil moisture retrieval in the third pole environment

The brightness temperature (TBp) observed by the Soil Moisture Active Passive (SMAP) satellite mission is significantly affected by the soil permittivity (εs), surface roughness and vegetation opacity (τp). This study assesses the impact of these factors on simulating the SMAP horizontally (p = H) and vertically (p = V) polarized TBp measurements and retrieving the liquid soil water content (θliq) for both frozen and thawed soils in the typical Tibetan desert and meadow ecosystems. For this investigation, the zero-order approximation of the radiative transfer equations, i.e., τ-ω emission model, is configured with surface roughness and τp parameterizations adopted by current SMAP soil moisture retrieval algorithms, and the εs is computed with the four-phase dielectric mixing model that is applicable for both frozen and thawed soils.For the Tibetan desert site, the τ-ω emission model with above configurations underestimates year-round the SMAP TBH measurements (bias > 20 K), while TBV are underestimated during the cold season. Implementation of a new surface roughness parameterization reduces the TBH underestimation, and the improved TBp simulations lead to better θliq retrievals produced by the single channel algorithm (SCA) using the TBV as well as TBH measurements. The remaining TBH and TBV underestimations are removed by further adopting a new εs parameterization.For the Tibetan meadow site, the τ-ω emission model overestimates both TBH and TBV during the warm season and underestimates TBH during the cold season when the vegetation is sparse. Implementation of the new surface roughness parameterization reduces the TBH underestimation, and further the TBp overestimation is mitigated by adopting a new τp parameterization derived from a discrete radiative transfer model previously developed and tested for the same site. The in-situ measured θliq dynamics are better captured by corresponding retrievals for both frozen and thawed soils with implementation of the new surface roughness and τp parameterizations, which reduces the unbiased RMSEs by more than 40%. The parameterizations developed in this study are useful to provide consistent and reasonable TBp simulations and θliq retrievals over the Tibetan Plateau for both frozen and thawed soils based on both SMAP TBH and TBV measurements.

Evaluation of the Tau–Omega Model for Passive Microwave Soil Moisture Retrieval Using SMAPEx Datasets

The parameters used for passive soil moisture retrieval algorithms reported in the literature encompass a wide range, leading to a large uncertainty in the applicability of those values. This paper presents an evaluation of the proposed parameterizations of the tau–omega model from 1) the soil moisture active passive (SMAP) algorithm theoretical basis document (ATBD) for global condition and 2) calibrated parameters from the National Airborne Field Experiment (NAFE’05) for Australian conditions, with special focus on the vegetation parameter b and roughness parameter $H_{R}$ . This study uses airborne L -band data and field observations from the SMAP experiments conducted in south-eastern Australia. Results show that the accuracy with the proposed parameterizations from SMAP ATBD was satisfactory at 100-m spatial resolution for maize (0.07 m3/m3) and pasture (0.07 m3/m3 ), while it decreased to 0.19 m3/m3 for wheat. Calibrated parameters from the NAFE’05 did not provide better results, with the accuracy of wheat degrading to 0.23 m3/m3. After a comprehensive site-specific calibration and validation at 100-m spatial resolution, this result was improved to 0.10 m3/m3. Further calibration and validation were performed at 1-km resolution against intensive ground sampling and at 3-km against in situ monitoring stations. Results showed an accuracy over grassland and cropland of 0.04 m3/m3 and 0.05 m3/m3 , respectively. This study also suggests that the parameters from SMAP ATBD show an underestimation of soil moisture, with the roughness parameter $H_{R}$ being too low for south-eastern Australian condition. Therefore, a new set of b and $H_{R}$ parameters for ten different land cover types was proposed in this study.

Comparison of soil moisture retrieval algorithms based on the synergy between SMAP and SMOS-IC

This study was carried out to evaluate possible improvements of the soil moisture (SM) retrievals from the SMAP observations, based on the synergy between SMAP and SMOS. We assessed the impacts of the vegetation and soil roughness parameters on SM retrievals from SMAP observations. To do so, the effects of three key input parameters including the vegetation optical depth (VOD), effective scattering albedo (ω) and soil roughness (HR) parameters were assessed with the emphasis on the synergy with the VOD product derived from SMOS-IC, a new and simpler version of the SMOS algorithm, over two years of data (April 2015 to April 2017). First, a comprehensive comparison of seven SM retrieval algorithms was made to find the best one for SM retrievals from the SMAP observations. All results were evaluated against in situ measurements over 548 stations from the International Soil Moisture Network (ISMN) in terms of four statistical metrics: correlation coefficient (R), root mean square error (RMSE), bias and unbiased RMSE (UbRMSE). The comparison of seven SM retrieval algorithms showed that the dual channel algorithm based on the additional use of the SMOS-IC VOD product (selected algorithm) led to the best results of SM retrievals over 378, 399, 330 and 271 stations (out of a total of 548 stations) in terms of R, RMSE, UbRMSE and both R & UbRMSE, respectively. Moreover, comparing the measured and retrieved SM values showed that this synergy approach led to an increase in median R value from 0.6 to 0.65 and a decrease in median UbRMSE from 0.09 m3/m3 to 0.06 m3/m3. Second, using the algorithm selected in a first step and defined above, the ω and HR parameters were calibrated over 218 rather homogenous ISMN stations. 72 combinations of various values of ω and HR were used for the calibration over different land cover classes. In this calibration process, the optimal values of ω and HR were found for the different land cover classes. The obtained results indicated that the impact of the VOD parameter on SM retrievals is more considerable than the effects of HR and ω. Overall, the inclusion of the VOD parameter in the SMAP SM retrieval algorithm was found to be a very interesting approach and showed the large potential benefit of the synergy between SMAP and SMOS.

Incorporation of Stem Water Content into Vegetation Optical Depth for Crops and Woodlands

Estimation of vegetation water content (VWC) by optical remote sensing improves soil moisture retrievals from passive microwave radiometry. For a variety of vegetation types, the largest unknown for predicting VWC is stem water content, which is assumed to be allometrically related to the water content of the plant canopy. For maize and soybean, measured stem water contents were highly correlated to canopy water contents, so VWC was calculated directly from the normalized difference infrared index (NDII), which contrasts scattering at near-infrared wavelengths with absorption of shortwave infrared wavelengths by liquid water. Woodland tree height is linearly related to woody stem volume, and hence to stem water content. We hypothesized that tree height is positively correlated with canopy water content, and thus with NDII. Airborne color-infrared imagery was acquired at two study areas in a mixed agricultural and woodland landscape, and photogrammetric structure-from-motion point clouds were derived to estimate tree heights. However, estimated tree heights were only weakly correlated with measured data acquired for validation. NDII was calculated from Landsat 8 Operational Line Imager (30-m pixel) and WorldView-3 (7.5 m pixel); but contrary to the hypothesis, NDII was not correlated with woodland tree height. Lastly, the interaction of woodland and crops stem water contents on total VWC in a mixed landscape were simulated for 2 days, one in the early summer and one in the late summer. VWC for the region varied from 2.5 to 3.0 kg m−2, which was just under a threshold for accuracy for soil moisture retrievals using Coriolis WindSat. Woodland tree height should be included as an ancillary data set along with land cover classification for soil moisture retrieval algorithms.