Soil moisture is the basic condition required for crop growth and development. Gaofen-3 (GF-3) is the first C-band synthetic-aperture radar (SAR) satellite of China, offering broad land and ocean imaging applications, including soil moisture monitoring. This study developed an approach to estimate soil moisture in agricultural areas from GF-3 data. An inversion technique based on an artificial neural network (ANN) is introduced. The neural network was trained and tested on a training sample dataset generated from the Advanced Integral Equation Model. Incidence angle and HH or VV polarization data were used as input variables of the ANN, with soil moisture content (SMC) and surface roughness as the output variables. The backscattering contribution from the vegetation was eliminated using the water cloud model (WCM). The acquired soil backscattering coefficients of GF-3 and in situ measurement data were used to validate the SMC estimation algorithm, which achieved satisfactory results (
Soil moisture content (SMC) is an important parameter in hydrological, biological, agricultural, and other processes [
Radar remote sensing is sensitive to soil moisture, due to the differences in the dielectric constants of soil and water; thus, the dielectric constant is one of the most important factors in the radar backscattering coefficient [
The above-mentioned models are commonly applied to bare soil and cannot be applied directly in vegetation cover areas, due to the multiple scattering effects of vegetation canopies [
The estimation of soil moisture is usually a nonlinear, ill-posed, complex process [
Gaofen-3 (GF-3) is the first Chinese civil C-band SAR. To date, there have been few soil moisture estimation methods developed for the GF-3 satellite. Therefore, we create an inversion technique based on ANN to estimate SMC over agricultural areas by combining GF-3 and Landsat-8 satellite data. The WCM was first applied to eliminate the effects of vegetation and to obtain the backscattering coefficients of bare soil. Then, the ANN was trained using a sample dataset generated form the AIEM. Meanwhile, field SMC data in an agricultural region with wheat as the main crop type were used to evaluate the potential of the GF-3 sensor for retrieving SMC data.
The remainder of this paper is organized as follows: Section
A study site located in the Luancheng County of Shijiazhuang city (centered at 114.65°E and 37.88°N; Figure
Left image shows the geolocation of the Luancheng study area in Shijiazhuang city, and the right image is Gaofen-3 (GF-3) Quad-Polarization Strip I (QPSI) data acquired on 27 May 2017. The red patches are residential areas, and the green triangles indicate the field measurement sampling points (point separation: 150 m) in each field.
GF-3 SAR and Landsat-8 OLI images were used in this study. In addition, 38 agriculture fields were selected to conduct in situ measurements, including those related to both soil and vegetation characteristics.
Microwave and optical satellite data were fused to reduce the effects of vegetation cover on the backscattering of soil moisture. Satellite data are listed in Table
Satellite data characteristics used in this study.
Type | Sensor | Spatial resolution (m) | Date (dd/mm/yy) | Retrieval purpose |
---|---|---|---|---|
Microwave | GF-3 | 8 | 27 May 2017 | Soil moisture |
Optical | Landsat-8 | 30 | 30 May 2017 | Vegetation water content |
In this study, one GF-3 SAR product acquired in the Quad-Polarization Strip I (QPSI) mode, processed up to level-1A SLC, was collected on May 27, 2017. Details of the QPSI model are listed in Table
Main technical specifications of each imaging mode.
Imaging mode | Incidence angle (°) | Resolution (m) | Imaging bandwidth (km) | Polarization mode | |||
---|---|---|---|---|---|---|---|
Nominal | Azimuth | Range | Nominal | Size | |||
Quad-Polarization Strip I | 20–41 | 8 | 8 | 6∼9 | 30 | 20∼35 | Full polarization |
The NASA’s Landsat-8 satellite carries two instruments: the Optical Land Imager (OLI) sensor and the Thermal Infrared Sensor (TIRS). It images the land surface using 11 spectral bands in the optical and thermal infrared domains with a spatial resolution of 30 to 100 m and a temporal resolution of 16 days [
We directly downloaded the land surface reflectance product, in which the image had been preprocessed; preprocessing included radiation calibration and atmospheric correction. The reflectance values of near-infrared (NIR) and short-wave infrared (SWIR) bands were used to estimate the normalized difference water index (NDWI) and VWC. Finally, Landsat-8 reflectance data were extracted from the sample points and combined with field measurements to build the relationship between vegetation water content (VWC) and NDWI.
Coincident with the GF-3 and Landsat-8 satellite overpasses, field campaign measurements of soil moisture and roughness, as well as crop biophysical parameters, were conducted over the 38 wheat fields. For each field, three sampling points were randomly selected (point separation 150 m). Soil volumetric moisture was measured using the oven-drying method (wet weight-dry weight) at a depth of 0–5 cm, given that the C-band radar signal is most sensitive to surface soil moisture. For each sampling point, measurements were collected at three locations that were uniformly distributed over 8 m (one pixel of the GF-3 satellite).The average soil moisture value of the three locations was considered the soil moisture of the sampling point. The VWC collected from 1.0 × 1.0 m squares selected at random was determined by weighting before and after oven-drying. Soil roughness was measured with a 1 m pin plate, including the root mean square height (
Our approach for the soil moisture estimation uses an ANN technique that combines GF-3 and Landsat-8 satellite data (Figure
Flow chart for soil moisture content (SMC) estimation using an artificial neural network (ANN).
VWC (kg/m2) is one of the most important parameters for the successful retrieval of SMC from microwave remote sensing observations [
The NDWI is a widely used and reliable indicator to assess the vegetation water status, which is sensitive to changes in VWC [
VWC was measured in 38 fields at the first sampling point. The above ground biomass was removed, and fresh and dry weights were used to compute the VWC. The relationship between VWC and NDWI was generated based on the least-squares fitting method, as follows:
The WCM, introduced by Attema and Ulaby [
Bare soil backscattering depends on the dielectric constant and surface roughness, as well as the SAR instrumental parameters [
The Dobson dielectric model is commonly used to describe the relationship between the effective dielectric constant of soil and soil moisture [
The incidence angle ranged from 20° to 60° with an interval of 1°. The
ANNs can mimic human learning and can build multivariate nonlinear relationships; as such, they have been widely used for estimating land surface parameters from remote sensing data [
The incidence angle and backscattering coefficient (HH or VV) were the input variables; the corresponding SMC and surface roughness were the output variables. The ANN was trained for HH and VV polarizations separately. One hidden layer and 30 neurons provided accurate SMC estimation within a reasonable computing time, by adding or removing these components from the model for both HH and VV polarizations. Therefore, the optimal ANN architecture (Figure
ANN architecture used for SMC estimation.
GF-3 satellite data were preprocessed to obtain backscattering coefficients of the agricultural area. The backscattering coefficients of soil were generated based on the WCM to eliminate the backscattering contribution of vegetation. The SMC can be estimated using the trained ANN and the backscattering coefficients of soil as the input parameter. Direct comparison of in situ SMC measurements with SMC estimations using GF-3 satellite data is a reliable way to assess the accuracy of the proposed SMC estimation algorithm. The precision and accuracy of SMC were estimated using two statistical indices: the
Information about VWC is an important parameter of the WCM, which is useful for retrieving soil moisture from GF-3 satellite data. NDWI was chosen to generate relationships with VWC based on Landsat-8 OLI land surface reflectance data and ground-based VWC measurements. Then, coefficients (
Vegetation water content image based on Landsat-8 data.
A sensitivity analysis between the GF-3 total radar backscatter
Correlations between in situ SMC and GF-3 radar backscatter
Linear fitting equations for in situ SMC and GF-3 radar backscattering coefficient
Polarization | Equation |
|
---|---|---|
HH |
|
0.146 |
VV |
|
0.120 |
The bare soil backscattering coefficient
Vegetation parameters in the water cloud model.
Polarization | Parameters | |
---|---|---|
|
|
|
HH | 0.0023 | 0.142 |
VV | 0.0019 | 0.127 |
The backscattering coefficients of 19 wheat fields (57 points) were extracted to analyze changes in
Changes in the backscattering coefficient (a) before and (b) after the elimination of vegetation information.
SMC estimation was achieved using an ANN method, in which the training sample dataset was generated based on the AIEM. Using the trained ANN, GF-3 satellite-measured soil backscatter data were used as the input and SMC as the output. To verify the reliability of the training sample dataset, the consistency between the simulated backscattering data (
Scatterplot of AIEM-simulated backscattering data (
The soil moisture estimation results are shown in Figure
Soil moisture retrieval results of (a) HH and (b) VV polarizations using GF-3 data.
In summary, the differences between the measured soil moisture and estimated soil moisture from GF-3 data were small, and the soil moisture retrieval results were satisfactory. These results indicate that the proposed soil moisture method for GF-3 data is reliable and that GF-3 data could achieve acceptable performance for soil moisture estimation. Thus, this approach shows the potential for providing the high-resolution soil moisture dataset for agricultural application, such as farmland soil moisture monitoring.
(a, b) Scatter diagram between in situ soil moisture and retrieved soil moisture.
We propose a soil moisture retrieval algorithm for agricultural regions that uses GF-3 satellite and Landsat-8 data based on the ANN method. The ANN structure was trained under a large range of land surface parameters, allowing the algorithm to have better adaptability to a variety of underlying conditions. The retrieval results using field soil moisture measurements obtained from an agricultural region with wheat as the dominated crop type showed that the proposed algorithm achieved satisfactory soil moisture estimation accuracy (e.g., RMSE = 0.042). The results indicated GF-3 satellite data had good performance on soil moisture retrieval, and the algorithm had potential to operationally estimate soil moisture from GF-3 satellite data. The major conclusions of this study are summarized as follows: VWC is an important factor for accurate retrieval of soil moisture under the vegetation cover. In this study, we combined microwave and optical remote sensing data to eliminate the contribution of backscattering coefficients caused by the vegetation. Landsat-8 OLI land surface reflectance data were chosen to complete the VWC estimation. The remotely sensed NDWI was used to generate the relationship with VWC through ground-based observation data. An AIEM-Dobson model was built to simulate the training sample dataset based on in situ measurements and GF-3 satellite parameters. This dataset includes the incidence angle, backscattering coefficients, and its corresponding SMC and surface roughness. The WCM was used to calculate the bare soil backscattering coefficient The backscattering coefficients of GF-3 satellite data were attenuated to different degrees compared to the total backscattering coefficients after the use of the WCM. Due to the different VWCs of each point, the attenuation degree is also different. The effect of vegetation contribution is large and must be removed before the soil moisture retrieval process; otherwise, it will influence the accuracy of soil moisture retrieval. The optimal ANN architecture in this study was determined as a three-layer network consisting of an input layer (three neurons: incidence angle and HH or VV backscattering), one hidden layer (30 neurons), and a two-component output layer (SMC and surface roughness). Our results and model sensitivity highlight the contribution of combined GF-3 SAR and Landsat-8 images using an ANN method for improving SMC estimates. HH polarization showed better SMC estimation performance than VV polarization.
Although satisfactory soil moisture retrieval performance was achieved, there were also several limitations to this study. The field measurement had many uncertainties that affected assessing soil moisture retrieval algorithms using remote sensing data. Real soil moisture values from ground-based measurements are difficult to match with pixel-level soil moisture estimates, although averaging multipoint measurements could reduce this error to a certain extent. Further work should focus on validating the proposed soil moisture retrieval algorithm using field soil moisture measurements with less uncertainty under various land cover-type conditions.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors declare no conflicts of interest.
Linlin Zhang, Qingyan Meng, and Qiuxia Xie conceived and designed the experiments. Linlin Zhang and Shun Yao performed the experiments and wrote the paper. Qingyan Meng, Xu Chen, and Ying Zhang contributed to paper revisions.
The authors would like to thank the National Key Research and Development Program (China’s 13th Five-Year Plan) (2017YFB0503900 and 2017YFB0503905), the Hainan Province Key Research and Development Plan (ZDYF2018231), and the Hainan Province Natural Science Foundation (417218 and 417219).