Soil moisture retrieval using SMOS brightness temperatures and a neural network trained on in situ measurements

2017 ◽  
Author(s):  
Nemesio J. Rodriguez-Fernandez ◽  
Veronica de Souza ◽  
Yann H. Kerr ◽  
Philippe Richaume ◽  
Ahmad Al Bitar
Keyword(s):  
Neural Network ◽  
Soil Moisture ◽  
In Situ Measurements ◽  
Brightness Temperatures

scholarly journals Comparing the Performance of Neural Network and Deep Convolutional Neural Network in Estimating Soil Moisture from Satellite Observations

2018 ◽  
Vol 10 (9) ◽  
pp. 1327 ◽  
Author(s):  
Lingling Ge ◽  
Renlong Hang ◽  
Yi Liu ◽  
Qingshan Liu
Keyword(s):  
Neural Network ◽  
Soil Moisture ◽  
Convolutional Neural Network ◽  
Hydrological Cycle ◽  
In Situ Measurements ◽  
Learning Phase ◽  
Satellite Observations ◽  
Deep Convolutional Neural Network ◽  
Testing Phase

Soil moisture (SM) plays an important role in hydrological cycle and weather forecasting. Satellite provides the only viable approach to regularly observe large-scale SM dynamics. Conventionally, SM is estimated from satellite observations based on the radiative transfer theory. Recent studies have demonstrated that the neural network (NN) method can retrieve SM with comparable accuracy as conventional methods. Here, we are interested in whether the NN model with more complex structures, namely deep convolutional neural network (DCNN), can bring about further improvement in SM retrievals when compared with the NN model used in recent studies. To achieve this objective, the same input data are used for the DCNN and NN models, including L-band Soil Moisture and Ocean Salinity (SMOS) brightness temperature (TB), C-band Advanced Scatterometer (ASCAT) backscattering coefficients, Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) and soil temperature. The target SM used to train the DCNN and NN models is the European Center for Medium-range Weather Forecasts Re-Analysis Interim (ERA-Interim) product. The experiment consists of two phases: the learning phase from 1 January to 31 December 2015 and the testing phase from 1 January to 31 December 2016. In the learning phase, we train the DCNN and NN models using the ERA-Interim SM. When evaluation between DCNN and NN against in situ measurements in the testing phase, we find that the temporal correlations between DCNN SM and in situ measurements are higher than those between NN SM and in situ measurements by 6 . 2 % and 2 . 5 % on ascending and descending orbits, respectively. In addition, from the perspective of temporal and spatial dynamics, the simulated SM values by DCNN/NN and the ERA-Interim SM agree relatively well at a global scale. Results suggest that both NN and DCNN models are effective in estimating SM from satellite observations, and DCNN can achieve slightly better performance than NN.

scholarly journals Sensitive Feature Evaluation for Soil Moisture Retrieval Based on Multi-Source Remote Sensing Data with Few In-Situ Measurements: A Case Study of the Continental U.S.

Water ◽  
2021 ◽  
Vol 13 (15) ◽  
pp. 2003
Author(s):  
Ling Zhang ◽  
Zixuan Zhang ◽  
Zhaohui Xue ◽  
Hao Li
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Feature Selection ◽  
Soil Moisture ◽  
Remote Sensing Data ◽  
In Situ Measurements ◽  
Feature Subset ◽  
Selection Methods ◽  
Sensing Data

Soil moisture (SM) plays an important role for understanding Earth’s land and near-surface atmosphere interactions. Existing studies rarely considered using multi-source data and their sensitiveness to SM retrieval with few in-situ measurements. To solve this issue, we designed a SM retrieval method (Multi-MDA-RF) using random forest (RF) based on 29 features derived from passive microwave remote sensing data, optical remote sensing data, land surface models (LSMs), and other auxiliary data. To evaluate the importance of different features to SM retrieval, we first compared 10 filter or embedded type feature selection methods with sequential forward selection (SFS). Then, RF was employed to establish a nonlinear relationship between the in-situ SM measurements from sparse network stations and the optimal feature subset. The experiments were conducted in the continental U.S. (CONUS) using in-situ measurements during August 2015, with only 5225 training samples covering the selected feature subset. The experimental results show that mean decrease accuracy (MDA) is better than other feature selection methods, and Multi-MDA-RF outperforms the back-propagation neural network (BPNN) and generalized regression neural network (GRNN), with the R and unbiased root-mean-square error (ubRMSE) values being 0.93 and 0.032 cm3/cm3, respectively. In comparison with other SM products, Multi-MDA-RF is more accurate and can well capture the SM spatial dynamics.

scholarly journals Mapping High Spatiotemporal-Resolution Soil Moisture by Upscaling Sparse Ground-Based Observations Using a Bayesian Linear Regression Method for Comparison with Microwave Remotely Sensed Soil Moisture Products

2021 ◽  
Vol 13 (2) ◽  
pp. 228
Author(s):  
Jian Kang ◽  
Rui Jin ◽  
Xin Li ◽  
Yang Zhang
Keyword(s):  
Soil Moisture ◽  
Large Scale ◽  
Spatial Scales ◽  
Microwave Remote Sensing ◽  
In Situ Measurements ◽  
Spatial Density ◽  
Linear Regression Method ◽  
Spatiotemporal Resolution ◽  
Pixel Scale

In recent decades, microwave remote sensing (RS) has been used to measure soil moisture (SM). Long-term and large-scale RS SM datasets derived from various microwave sensors have been used in environmental fields. Understanding the accuracies of RS SM products is essential for their proper applications. However, due to the mismatched spatial scale between the ground-based and RS observations, the truth at the pixel scale may not be accurately represented by ground-based observations, especially when the spatial density of in situ measurements is low. Because ground-based observations are often sparsely distributed, temporal upscaling was adopted to transform a few in situ measurements into SM values at a pixel scale of 1 km by introducing the temperature vegetation dryness index (TVDI) related to SM. The upscaled SM showed high consistency with in situ SM observations and could accurately capture rainfall events. The upscaled SM was considered as the reference data to evaluate RS SM products at different spatial scales. In regard to the validation results, in addition to the correlation coefficient (R) of the Soil Moisture Active Passive (SMAP) SM being slightly lower than that of the Climate Change Initiative (CCI) SM, SMAP had the best performance in terms of the root-mean-square error (RMSE), unbiased RMSE and bias, followed by the CCI. The Soil Moisture and Ocean Salinity (SMOS) products were in worse agreement with the upscaled SM and were inferior to the R value of the X-band SM of the Advanced Microwave Scanning Radiometer 2 (AMSR2). In conclusion, in the study area, the SMAP and CCI SM are more reliable, although both products were underestimated by 0.060 cm3 cm−3 and 0.077 cm3 cm−3, respectively. If the biases are corrected, then the improved SMAP with an RMSE of 0.043 cm3 cm−3 and the CCI with an RMSE of 0.039 cm3 cm−3 will hopefully reach the application requirement for an accuracy with an RMSE less than 0.040 cm3 cm−3.

scholarly journals SMOS Neural Network Soil Moisture Data Assimilation in a Land Surface Model and Atmospheric Impact

2019 ◽  
Author(s):  
Nemesio Rodriguez-Fernandez ◽  
Patricia de Rosnay ◽  
Clement Albergel ◽  
Philippe Richaume ◽  
Filipe Aires ◽  
...  
Keyword(s):  
Neural Network ◽  
Soil Moisture ◽  
Data Assimilation ◽  
Land Surface ◽  
Positive Impact ◽  
Land Surface Model ◽  
Surface Model ◽  
Northern Asia ◽  
The Impact

The assimilation of Soil Moisture and Ocean Salinity (SMOS) data into the ECMWF (European Centre for Medium Range Weather Forecasts) H-TESSEL (Hydrology revised - Tiled ECMWF Scheme for Surface Exchanges over Land) model is presented. SMOS soil moisture (SM) estimates have been produced specifically by training a neural network with SMOS brightness temperatures as input and H-TESSEL model SM simulations as reference. This can help the assimilation of SMOS information in several ways: (1) the neural network soil moisture (NNSM) data have a similar climatology to the model, (2) no global bias is present with respect to the model even if regional differences can exist. Experiments performing joint data assimilation (DA) of NNSM, 2 metre air temperature and relative humidity or NNSM-only DA are discussed. The resulting SM was evaluated against a large number of in situ measurements of SM obtaining similar results to those of the model with no assimilation, even if significant differences were found from site to site. In addition, atmospheric forecasts initialized with H-TESSEL runs (without DA) or with the analysed SM were compared to measure of the impact of the satellite information. Although, NNSM DA has an overall neutral impact in the forecast in the Tropics, a significant positive impact was found in other areas and periods, especially in regions with limited in situ information. The joint NNSM, T2m and RH2m DA improves the forecast for all the seasons in the Southern Hemisphere. The impact is mostly due to T2m and RH2m, but SMOS NN DA alone also improves the forecast in July- September. In the Northern Hemisphere, the joint NNSM, T2m and RH2m DA improves the forecast in April-September, while NNSM alone has a significant positive effect in July-September. Furthermore, forecasting skill maps show that SMOS NNSM improves the forecast in North America and in Northern Asia for up to 72 hours lead time.

Towards an operationally capable satellite-based 0-100 cm soil moisture dataset from C3S

2021 ◽  
Author(s):  
Adam Pasik ◽  
Wolfgang Preimesberger ◽  
Bernhard Bauer-Marschallinger ◽  
Wouter Dorigo
Keyword(s):  
Soil Moisture ◽  
Soil Profile ◽  
Land Surface ◽  
Root Zone ◽  
Soil Layer ◽  
Land Surface Model ◽  
In Situ Measurements ◽  
Surface Model ◽  
T Values

<p>Multiple satellite-based global surface soil moisture (SSM) datasets are presently available, these however, address exclusively the top layer of the soil (0-5cm). Meanwhile, root-zone soil moisture cannot be directly quantified with remote sensing but can be estimated from SSM using a land surface model. Alternatively, soil water index (SWI; calculated from SSM as a function of time needed for infiltration) can be used as a simple approximation of root-zone conditions. SWI is a proxy for deeper layers of the soil profile which control evapotranspiration, and is hence especially important for studying hydrological processes over vegetation-covered areas and meteorological modelling.</p><p>Here we introduce the advances in our work on the first operationally capable SWI-based root-zone soil moisture dataset from C3S Soil Moisture v201912 COMBINED product, spanning the period 2002-2020. The uniqueness of this dataset lies in the fact that T-values (temporal lengths ruling the infiltration) characteristic of SWI were translated into particular soil depths making it much more intuitive, user-friendly and easily applicable. Available are volumetric soil moisture values for the top 1 m of the soil profile at 10 cm intervals, where the optimal T-value (T-best) for each soil layer is selected based on a range of correlation metrics with in situ measurements from the International Soil Moisture Network (ISMN) and the relevant soil and climatic parameters.<br>Additionally we present the results of an extensive global validation against in situ measurements (ISMN) as well as the results of investigations into the relationship between a range of soil and climate characteristics and the optimal T-values for particular soil depths.</p>

scholarly journals Accuracy and Transferability of Artificial Neural Networks in Predicting in Situ Root-Zone Soil Moisture for Various Regions across the Globe

Water ◽  
2020 ◽  
Vol 12 (11) ◽  
pp. 3109
Author(s):  
Roïya Souissi ◽  
Ahmad Al Bitar ◽  
Mehrez Zribi
Keyword(s):  
Neural Network ◽  
Soil Moisture ◽  
Soil Texture ◽  
Surface Soil ◽  
Root Zone ◽  
Data Filtering ◽  
Surface Soil Moisture ◽  
The Mean ◽  
Artificial Neural

This paper explores the accuracy in using an artificial neural network (ANN) to estimate root-zone soil moisture (RZSM) at multiple worldwide locations using only in situ surface soil moisture (SSM) as a training dataset. The paper also addresses the transferability of the trained ANN across climatic and soil texture conditions. Data from the International Soil Moisture Network (ISMN) were collected for several networks with variable soil texture and climate classes. Several scaling, feature extraction, and training approaches were tested. An artificial neural network employing rolling averages (ANNRAV) of SSM over 10, 30, and 90 days was developed. The results show that applying a standard scaling (SSCA) to the ANN input features improves the correlation, Nash–Sutcliffe efficiency (NSE), and root mean square error (RMSE) for 52%, 91%, and 87%, respectively, of the tested stations, compared to MinMax scaling (MMSCA). Different training sets are suggested, namely, training on data from all networks, data from one network, or data of all networks excluding one. Based on these trainings, new transferability (TranI) and contribution (ContI) indices are defined. The results show that one network cannot provide the best prediction accuracy if used alone to train the ANN. They also show that the removal of the less contributing networks enhances performance. For example, elimination of the densest network (SCAN) from the training enhances the mean correlation by 20.5% and the mean NSE by 42.5%. This motivates the implementation of a data filtering technique based on the ANN’s performance. A median, max, and min correlation of 0.77, 0.96, and 0.65, respectively, are obtained by the model after data filtering. The performances are also analyzed with respect to the covered climatic regions and soil texture, providing insights into the robustness and limitations of the approach, namely, the need for complementary information in highly evaporative regions. In fact, the ANN using only SSM to predict RZSM has low performance when decoupling between the surface and root zones is observed. The application of ANN to obtain spatialized RZSM will require integrating remote sensing-based surface soil moisture in the future.

Comparing Assimilation of Soil Moisture and C-band Backscatter in High Resolution Land Surface Model

2020 ◽  
Author(s):  
Leqiang Sun ◽  
Stéphane Belair ◽  
Marco Carrera ◽  
Bernard Bilodeau
Keyword(s):  
Soil Moisture ◽  
Land Surface ◽  
Land Surface Model ◽  
Open Loop ◽  
In Situ Measurements ◽  
Lookup Table ◽  
Surface Model ◽  
Loop Model ◽  
The Impact

<p>Canadian Space Agency (CSA) has recently started receiving and processing the images from the recently launched C-band RADARSAT Constellation Mission (RCM). The backscatter and soil moisture retrievals products from the previously launched RADARSAT-2 agree well with both in-situ measurements and surface soil moisture modeled with land surface model Soil, Vegetation, and Snow (SVS). RCM will provide those products at an even better spatial coverage and temporal resolution. In preparation of the potential operational application of RCM products in Canadian Meteorological Center (CMC), this paper presents the scenarios of assimilating either soil moisture retrieval or outright backscatter signal in a 100-meter resolution version of the Canadian Land Data Assimilation System (CaLDAS) on field scale with time interval of three hours. The soil moisture retrieval map was synthesized by extrapolating the regression relationship between in-situ measurements and open loop model output based on soil texture lookup table. Based on this, the backscatter map was then generated with the surface roughness retrieved from RADARSAT-2 images using a modified Integral Equation Model (IEM) model. Bias correction was applied to the Ensemble Kalman filter (EnKF) to mitigate the impact of nonlinear errors introduced by multi-sourced perturbations. Initial results show that the assimilation of backscatter is as effective as assimilating soil moisture retrievals. Compared to open loop, both can improve the analysis of surface moisture, particularly in terms of reducing bias.  </p>

Sign in / Sign up

Export Citation Format

Share Document