Some critical remarks on the use of C-band radar data for soil moisture detection

Abstract

Multifrequency microwave backscatter from soils under different agricultural crops and different moisture conditions was measured during the LOTREX campaign (Land Surface Transverse Experiment. 26 June-21 July, 1989) in northern West Germany (LOTREX is part of the International Satellite Land-Surface Climatology Project (ISLSCP)). The data were gathered with an airborne coherent Doppler radar scatterometer at an off-nadir angle of 23^° as it was multiplexed through its L-, C-, X- and K_u-bahds. The frequency dependency of the backscatter power spectra was analysed and published elsewhere. In this Letter we discuss polarization effects in the C-band.     

Mapping surface roughness and soil moisture using multi-angle radar imagery without ancillary data

The Integral Equation Model (IEM) is the most widely-used, physically based radar backscatter model for sparsely vegetated landscapes. In general, IEM quantifies the magnitude of backscattering as a function of moisture content and surface roughness, which are unknown, and the known radar configurations. Estimating surface roughness or soil moisture by solving the IEM with two unknowns is a classic example of under-determination and is at the core of the problems associated with the use of radar imagery coupled with IEM-like models. This study offers a solution strategy to this problem by the use of multi-angle radar images, and thus provides estimates of roughness and soil moisture without the use of ancillary field data. Results showed that radar images can provide estimates of surface soil moisture at the watershed scale with good accuracy. Results at the field scale were less accurate, likely due to the influence of image speckle. Results also showed that subsurface roughness caused by rock fragments in the study sites caused error in conventional applications of IEM based on field measurements, but was minimized by using the multi-angle approach.     

A genetic algorithm to simultaneously retrieve land surface roughness and soil wetness

Two key parameters to affect microwave backscattering from the land surface are the surface roughness and soil wetness. A novel genetic algorithm is developed for multi-parameter retrieval of land surface roughness and soil wetness from angular backscattering observations. Parameters of wetness and roughness are encoded into genes. Genes are constituents of chromosomes, which undergo optimal selection based on a natural evolutionary process in the genetic algorithm. The theoretical model of a two-scale rough surface is employed for computation of the cost function. Results retrieved using this genetic algorithm are compared well with ground data measurements. This study presents an example of the genetic algorithm for application of multi-parameter retrieval in remote sensing.     

A new method to retrieve significant wave height from X-band marine radar image sequences

A new method is proposed to retrieve significant wave height (SWH) from X-band marine radar image sequences. To reduce the inhomogeneity of the nearshore wave field, the principal component (PC) of the radar image sequence is extracted by empirical orthogonal function (EOF) analysis. To measure the information contained in each PC, the Shannon entropy is introduced after the PC is normalized. Based on the information contained in the wave field, a linear relationship is established to retrieve the SWH from the Shannon entropy of the PC. The method is validated by comparison with measurements from in situ buoys: the root mean square error between the SWH measured by a buoy and the retrieved value is 0.22 m, while the corresponding bias and correlation coefficient are 0.01 m and 0.92, respectively. The physical meanings of different EOF modes decomposed from the wave field are also discussed.     

Analysis of TerraSAR-X data sensitivity to bare soil moisture, roughness, composition and soil crust

Soils play a key role in shaping the environment and in risk assessment. We characterized the soils of bare agricultural plots using TerraSAR-X (9.5 GHz) data acquired in 2009 and 2010. We analyzed the behavior of the TerraSAR-X signal for two configurations, HH-25° and HH-50°, with regard to several soil conditions: moisture content, surface roughness, soil composition and soil-surface structure (slaking crust).The TerraSAR-X signal was more sensitive to soil moisture at a low (25°) incidence angle than at a high incidence angle (50°). For high soil moisture (> 25%), the TerraSAR-X signal was more sensitive to soil roughness at a high incidence angle (50°) than at a low incidence angle (25°).The high spatial resolution of the TerraSAR-X data (1 m) enabled the soil composition and slaking crust to be analyzed at the within-plot scale based on the radar signal. The two loamy-soil categories that composed our training plots did not differ sufficiently in their percentages of sand and clay to be discriminated by the X-band radar signal.However, the spatial distribution of slaking crust could be detected when soil moisture variation is observed between soil crusted and soil without crust. Indeed, areas covered by slaking crust could have greater soil moisture and consequently a greater backscattering signal than soils without crust.     

Semi-empirical calibration of the IEM backscattering model using radar images and moisture and roughness field measurements

Estimating surface parameters by radar-image inversion requires the use of well-calibrated backscattering models. None of the existing models is capable of correctly simulating scatterometer or satellite radar data. We propose a semi-empirical calibration of the Integral Equation Model (IEM) backscattering model in order to better reproduce the radar backscattering coefficient over bare agricultural soils. As correlation length is not only the least accurate but also the most difficult to measure of the parameters required in the models, we propose that it be replaced by a calibration parameter that would be estimated empirically from experimental databases of radar images and field measurements. This calibration was carried out using a number of radar configurations with different incidence angles, polarization configurations, and radar frequencies. Using several databases, the relationship between the calibration parameter and the surface roughness was determined for each radar configuration. In addition, the effect of the correlation function shape on IEM performance was studied using the three correlation functions (exponential, fractal, and Gaussian). The calibrated version of the IEM was then validated using another independent set of experimental data. The results show good agreement between the backscattering coefficient provided by the radar systems and that simulated by the calibrated version of the IEM. This calibrated version of the IEM can be used in inversion procedures to retrieve surface roughness and/or moisture values from radar images.     

A MODIS-based perpendicular moisture index to retrieve leaf moisture content of forest canopies

Moisture dictates vegetation susceptibility to fire ignition and propagation. Various spectral indices have been proposed for the estimation of equivalent water thickness (EWT), which is defined as the mass of liquid water per unit of leaf surface. However, fire models use live fuel moisture content (LFMC) as a measure of vegetation moisture. LFMC is defined as the ratio of the mass of the liquid water in a leaf over the mass of dry matter, and traditional spectral indices are not as effective as with EWT in capturing LFMC variability. The aim of this research was to explore the potential of the Moderate Resolution Imaging Spectroradiometer (MODIS) on board Terra and Aqua satellites in retrieving LFMC from top of the canopy reflectance, and to develop a new spectral index sensitive to this parameter. All the analyses were based on synthetic canopy spectra constructed by coupling the PROSPECT (leaf optical properties model) and SAIL (Scattering by Arbitrarily Inclined Leaves) radiative transfer models. Simulated top of the canopy spectra were then convolved to MODIS ‘land’ channels 1–7 spectral response functions. All band pairs were evaluated to determine the subspace of MODIS measurements where the separability of points based on their value of LFMC was the highest. This led to the identification of isolines of LFMC in the plane defined by MODIS reflectance measurements in channels 2 and 5; the isolines are straight and parallel, and ordered from lower to higher values of LFMC. This observation allowed the construction of a novel spectral index that is directly related to LFMC – the perpendicular moisture index (PMI). This index measures the distance of a point in the plane spanned by reflectance measurements in MODIS channels 2 and 5 from a reference line, that of completely dry vegetation. Validation against simulated data showed that PMI exhibits a linear relationship with LFMC. When the vegetation cover is dense, the LFMC explains most of the variability in the PMI (R² = 0.70 when LAI > 2; R² = 0.87 when LAI > 4). When the LAI is lower, the contribution of soil background to the measured reflectance increases, and the index underestimates LFMC. The PMI was also validated against the LOPEX93 (Leaf Optical Properties Experiment 1993) data set of leaf optical and biophysical measurements, scaled to canopy reflectance with SAIL, showing acceptable results (R² = 0.56 when LAI > 2; R² = 0.63 when LAI > 4).     

Likelihood parameter estimation for calibrating a soil moisture model using radar bakscatter

Land surface model parameter estimation can be performed using soil moisture information provided by synthetic aperture radar imagery. The presence of speckle necessitates aggregating backscatter measurements over large (> 100 m × 100 m) land areas in order to derive reliable soil moisture information from imagery, and a model calibrated to such aggregated information can only provide estimates of soil moisture at spatial resolutions required for reliable speckle accounting. A method utilizing the likelihood formulation of a probabilistic speckle model as the calibration objective function is proposed which will allow for calibrating land surface models directly to radar backscatter intensity measurements in a way which simultaneously accounts for model parameter- and speckle-induced uncertainty. The method is demonstrated using the NOAH land surface model and Advanced Integral Equation Method (AIEM) backscatter model calibrated to SAR imagery of an area in the Southwestern United States, and validated against in situ soil moisture measurements. At spatial resolutions finer than 100 m × 100 m NOAH and AIEM calibrated using the proposed radar intensity likelihood parameter estimation algorithm predict surface level soil moisture to within 4% volumetric water content 95% of the time, which is an improvement over a 95% prediction confidence of 10% volumetric water content by the same models calibrated directly to soil moisture information derived from synthetic aperture radar imagery at the same scales. Results suggest that much of this improvement is due to increased ability to simultaneously estimate NOAH parameters and AIEM surface roughness parameters.     

A model for retrieving soil moisture saturation with Landsat remotely sensed data

Soil moisture saturation indicates the capability of the vegetation humus layer and the soil layer to reabsorb and drain water in an area; it is crucial in predicting natural disasters, such as landslides and droughts. In this article, a model was created to retrieve soil moisture saturation based on multispectral remotely sensed data. Soil brightness and soil wetness, calculated from the tasseled cap transformation, were utilized to obtain soil moisture saturation. With the above model, a soil moisture saturation map of Maoergai District, which is located on the upper Minjiang River in northern Sichuan Province in the south-west of China, was created from a Landsat Enhanced Thematic Mapper Plus (ETM+) image in July 2002. Then, the soil type data and the vegetation distribution data of the year 2000 were used to evaluate the model. The result shows that the model for soil moisture saturation is viable and that the vegetation type, vegetation distribution and soil type have strong correlation with soil moisture saturation.     

Investigating the potential for soil moisture and surface roughness monitoring in drylands using ERS SAR data

Multitemporal ERS-1 and ERS-2 SAR data were acquired for northern Jordan between 1995 and 1997 to investigate changes in the backscatter coefficients of a range of typical desert land surfaces. The changes in backscatter found were ascribed to variations in surface soil moisture, and changes in surface roughness caused by a range of natural and anthropogenic factors. Data collected from monitored sites were input into the Integral Equation Model (IEM). The model outputs were strongly correlated with observed backscatter coefficients (r ²=0.84). The results show that the successful monitoring of soil moisture in these environments is strongly dependent on the surface roughness. On surfaces with RMS height 0.5 cm, the sensitivity of the backscatter coefficient to changes in surface microtopography did not allow accurate soil moisture estimation. Microtopographic change on rougher surfaces has less influence on the backscatter coefficient, and the probability of soil moisture estimation from SAR imagery is greater. These results indicate that knowledge of the surface conditions (both in terms of surface roughness and geomorphology) is essential for accurate soil moisture monitoring, whether in a research or operational context. The potential benefits of these findings are discussed in the context of the Jordan Badia Research and Development Project.     

Sensitivity of bistatic scattering to soil moisture and surface roughness of bare soils

The sensitivity of bistatic scattering coefficient σ° to soil moisture content (SMC) and surface roughness was investigated by means of model simulations of the incoherent scattered fields performed with the advanced integral equation model (AIEM) and the second order small perturbation model (SPM). The study was performed by simulating scattering on the whole upper half space, for different values of incident angles. The achieved results, represented as maps of σ° as a function of azimuth and zenith angles, were evaluated by means of a quality index which takes into consideration the effect of roughness on SMC measurement. The sensitivity analysis has pointed out that for measuring SMC a bistatic observation, by itself or combined with the monostatic one, can make appreciable improvements with respect to classical monostatic radar. Appendix A contains the AIEM formulas corrected for several typographical errors present in the specific literature.     

A method to retrieve subpixel fire temperature and fire area using MODIS data

Methods for retrieving subpixel fire temperature and fire area have been developed over several years, but the retrieval accuracies of these methods require further improvement. In this study, a channel of the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor centred at 2.1 μm and associated with the MODIS 4.0 2.1 μm channel is used to retrieve the temperature and area of fires. To test the feasibility of using the 2.1 μm channel for retrieval, the fire contribution ratios of MODIS 2.1, 4.0 and 11.0 μm channels are first examined using simulated surface radiance. Considering the difficulties in obtaining real-time validation data and in evaluating the retrieval accuracies, simulated MODIS data are used for this study. A modified method, which combines MODIS 2.1 and 4.0 μm channels, is introduced and described in detail. Compared with the traditional method, which utilizes a combination of 4.0 and 11.0 μm channels (Dozier 1981 Dozier, J. 1981. A method for satellite identification of surface temperature fields of subpixel resolution. Remote Sensing of Environment, 11: 221–229. [Crossref], [Web of Science ®] , [Google Scholar]), the results show that the 2.1 μm channel is more sensitive to active fires and the large area of fires than the 11.0 μm channel, but is less sensitive to smouldering fires and small fires. The modified method that we propose has better performance and higher accuracy in active fires (temperature ≥ 800 K) and in large fires (area ≥ 0.5%). However, the traditional method is more accurate for smouldering fires and small fires. Finally, a sensitivity analysis is performed to estimate the uncertainty in assessing fire temperature and area. Experimental results indicate that under realistic conditions (fire temperatures of approximately 1000 K and a fire fractional area greater than 0.005), the retrieval errors for fire temperature and fire area are ±35 K and 20%, respectively.     

A methodology to retrieve text documents from multiple databases

This paper presents a methodology for finding the n most similar documents across multiple text databases for any given query and for any positive integer n. This methodology consists of two steps. First, the contents of databases are indicated approximately by database representatives. Databases are ranked using their representatives with respect to the given query. We provide a necessary and sufficient condition to rank the databases optimally. In order to satisfy this condition, we provide three estimation methods. One estimation method is intended for short queries; the other two are for all queries. Second, we provide an algorithm, OptDocRetrv, to retrieve documents from the databases according to their rank and in a particular way. We show that if the databases containing the n most similar documents for a given query are ranked ahead of other databases, our methodology will guarantee the retrieval of the n most similar documents for the query. When the number of databases is large, we propose to organize database representatives into a hierarchy and employ a best-search algorithm to search the hierarchy. It is shown that the effectiveness of the best-search algorithm is the same as that of evaluating the user query against all database representatives.     

Statistical properties of soil moisture images derived from Radarsat-1 SAR data

In this article, we report on the assessment of the spatial variability of soil moisture using synthetic aperture radar (SAR) data. The imagery was acquired during five different periods over the Roseau River watershed in southern Manitoba, Canada. For validation purposes, ground measurements were carried out at 62 locations simultaneous with the satellite data acquisitions. The first step in this analysis was to assess the performance of the Integral Equation Model (IEM) in simulating backscatter coefficients for selected bare soils. In order to reduce the surface roughness effect on radar backscatter, a semi-empirical calibration technique was implemented. This calibrated model was then implemented in a simplex inversion routine in order to estimate and map soil moisture. Derived spatial patterns of near-surface moisture content were then examined using scale analyses. It has been confirmed that the variance of radar-based soil moisture images follows power law decay versus the observation scale. Also, more explicit analysis of the same soil moisture maps shows a ln–ln linear spatial scale with statistical moments. Concave shape dependency of the corresponding slopes with the moment order was observed during all radar acquisition periods. The latter indicates the presence of multifractal effects.     

Appropriate scale of soil moisture retrieval from high resolution radar imagery for bare and minimally vegetated soils

This research investigates the appropriate scale for watershed averaged and site specific soil moisture retrieval from high resolution radar imagery. The first approach involved filtering backscatter for input to a retrieval model that was compared against field measures of soil moisture. The second approach involved spatially averaging raw and filtered imagery in an image-based statistical technique to determine the best scale for site-specific soil moisture retrieval. Field soil moisture was measured at 1225 m² sites in three watersheds commensurate with 7 m resolution Radarsat image acquisition. Analysis of speckle reducing block median filters indicated that 5 × 5 filter level was the optimum for watershed averaged estimates of soil moisture. However, median filtering alone did not provide acceptable accuracy for soil moisture retrieval on a site-specific basis. Therefore, spatial averaging of unfiltered and median filtered power values was used to generate backscatter estimates with known confidence for soil moisture retrieval. This combined approach of filtering and averaging was demonstrated at watersheds located in Arizona (AZ), Oklahoma (OK) and Georgia (GA). The optimum ground resolution for AZ, OK and GA study areas was 162 m, 310 m, and 1131 m respectively obtained with unfiltered imagery. This statistical approach does not rely on ground verification of soil moisture for validation and only requires a satellite image and average roughness parameters of the site. When applied at other locations, the resulting optimum ground resolution will depend on the spatial distribution of land surface features that affect radar backscatter. This work offers insight into the accuracy of soil moisture retrieval, and an operational approach to determine the optimal spatial resolution for the required application accuracy.     

Passive microwave sensing of soil moisture content: The effects of soil bulk density and surface roughness

Microwave radiometric measurements over bare fields of different surface roughness were made at frequencies of 1.4 GHz, 5 GHz, and 10.7 GHz to study the frequency dependence, as well as the possible time variation, of surface roughness. An increase in surface roughness was found to increase the brightness temperature af soils and reduce the slope of regression between brightness temperature and soil moisture content. The frequency dependence of the surface roughness effect was relatively weak when compared with that of the vegetation effect. Radiometric time-series observations over a given field indicate that field surface roughness might gradually diminish with time, especially after a rainfall or irrigation. The variation of surface roughness increases the uncertainty of remote soil moisture estimates by microwave radiometry. Three years of radiometric measurements over a test site revealed a possible inconsistency in the soil bulk density determination, which is an important factor in the interpretation of radiometric data.     

A neural-network model to retrieve CDOM absorption from in situ measured hyperspectral data in an optically complex lake: Lake Taihu case study

Coloured dissolved organic matter (CDOM) is an important water component that affects water colour and ecological environment under water. The remote estimation of CDOM is always a challenge in the field of water-colour remote sensing owing to its weak signal. To further study the CDOM-retrieval approach, field experiments, including water-quality analysis and spectral measurements, were carried out in Lake Taihu waters from 8 to 21 November 2007. On the foundation of analysing water-inherent optical properties, sensitive spectral factors were selected, and then neural-network models were established for retrieving CDOM. The results show that the model with 10 nodes in the hidden layer performs best, yielding a correlation coefficient (R) of 0.887 and a root-mean-square error of 0.156 m^?1. Meanwhile, the predictive errors of the model developed here and the previously proposed algorithms were compared with each other. The mean value of the relative error of the former is 12.8% (standard deviation of 29.9%), and is much lower than its counterpart of other models, which indicates that the developed model has a higher accuracy for CDOM retrieval in Lake Taihu waters. Meanwhile, other datasets collected at different times were also imported into the model for applicability analysis; the derived errors suggest a relatively good performance of the model. This research firstly explores the CDOM retrieval in optically complex lake waters, and the corresponding findings support a technical framework for accurately extracting CDOM information in Lake Taihu waters, based on an adequate understanding of water optical properties.     

Potential of Dubois model for soil moisture retrieval in prairie areas using SAR and optical data

Soil moisture retrieval is often confounded by the influence of vegetation and surface roughness on the backscattered radar signal in vegetated areas. In this study, a semi-empirical methodology is proposed to retrieve soil moisture in prairie areas. The effect of vegetation is eliminated by the ratio vegetation method and water cloud model (WCM), respectively. The conditions of vegetation are characterized by leaf area index (LAI), vegetation water content (VWC), normalized difference vegetation index (NDVI), and enhanced vegetation index (EVI), respectively. To remove the dependence on surface roughness, the dielectric constant is explicitly expressed as the function of co-polarization backscattering coefficients and sensor parameters based on the Dubois model. The ground measurements and satellite data collected from the Ruoergai and Wutumeiren prairies of China allow for validating the feasibility and effectiveness of the proposed methodology. From the perspective of soil moisture retrieval accuracy, the ratio vegetation method performs better than WCM. In the Ruoergai prairie, the best soil moisture retrieval result is obtained when EVI is used, with correlation coefficient (r) and root mean square error (RMSE) of 0.87 and 3.50 vol.%, respectively. While in the Wutumeiren prairie, the lowest retrieval error is obtained when LAI is used, with r and RMSE values of 0.79 and 5.73 vol.%, respectively. These results demonstrate that the Dubois model has a potential for enhancing soil moisture retrieval in prairie areas using synthetic aperture radar (SAR) and optical data.