Estimation of high-resolution soil moisture using machine learning, satellite observations and ground measurements. A case study in a hilly agricultural region

2020 ◽  
Author(s):  
Luca Zappa ◽  
Matthias Forkel ◽  
Angelika Xaver ◽  
Wouter Dorigo
Keyword(s):  
Soil Moisture ◽  
Random Forest ◽  
Spatial Resolution ◽  
Vegetation Cover ◽  
Temporal Dynamics ◽  
Remotely Sensed ◽  
Training Data ◽  
Future Research ◽  
Microwave Sensors

<p>Remotely sensed data from microwave sensors have been successfully used to retrieve soil moisture on a global scale. In particular, passive and active microwave sensors with large footprints can observe the same location with a (sub-)daily frequency, but typically are characterized by spatial resolutions in the order of tens of km. Therefore, such coarse scale products can accurately capture the temporal dynamics of soil moisture but are inadequate in providing spatial details. However, several agricultural and hydrological applications could greatly benefit from soil moisture observations with a sub-kilometer spatial resolution while preserving a daily revisit time.</p><p>Here, we present a framework for downscaling coarse resolution satellite soil moisture products (ASCAT and SMAP) to high spatial resolution. In particular, we build robust relationships between remotely sensed soil moisture and ancillary variables on soil texture, topography, and vegetation cover. Such relationship is built through Random Forest regressions, trained against in-situ measurements of soil moisture. The proposed approach is developed and tested in an agricultural catchment equipped with a high-density network of in-situ sensors. Our results show a strong consistency between the downscaled and the observed spatio-temporal patterns of soil moisture. Furthermore, including a proxy of vegetation cover in the Random Forest regressions results in considerable improvements of the downscaling performance. Finally, if only limited training data can be used, priority should be given to increase the number of sensor locations to adequately cover the spatial heterogeneity, rather than expanding the duration of the measurements. </p><p>Future research will focus on including additional ancillary variables as model predictors, e.g. Land Surface Temperature or backscatter, and on applying the downscaling framework to other regions with similar environmental and climatic conditions.</p>

scholarly journals New Downscaling Approach Using ESA CCI SM Products for Obtaining High Resolution Surface Soil Moisture

2020 ◽  
Vol 12 (7) ◽  
pp. 1119 ◽  
Author(s):  
Jovan Kovačević ◽  
Željko Cvijetinović ◽  
Nikola Stančić ◽  
Nenad Brodić ◽  
Dragan Mihajlović
Keyword(s):  
Soil Moisture ◽  
High Resolution ◽  
Random Forest ◽  
Spatial Resolution ◽  
Surface Soil ◽  
Initial Step ◽  
Surface Soil Moisture ◽  
In Situ Data ◽  
Random Forest Models

ESA CCI SM products have provided remotely-sensed surface soil moisture (SSM) content with the best spatial and temporal coverage thus far, although its output spatial resolution of 25 km is too coarse for many regional and local applications. The downscaling methodology presented in this paper improves ESA CCI SM spatial resolution to 1 km using two-step approach. The first step is used as a data engineering tool and its output is used as an input for the Random forest model in the second step. In addition to improvements in terms of spatial resolution, the approach also considers the problem of data gaps. The filling of these gaps is the initial step of the procedure, which in the end produces a continuous product in both temporal and spatial domains. The methodology uses combined active and passive ESA CCI SM products in addition to in situ soil moisture observations and the set of auxiliary downscaling predictors. The research tested several variants of Random forest models to determine the best combination of ESA CCI SM products. The conclusion is that synergic use of all ESA CCI SM products together with the auxiliary datasets in the downscaling procedure provides better results than using just one type of ESA CCI SM product alone. The methodology was applied for obtaining SSM maps for the area of California, USA during 2016. The accuracy of tested models was validated using five-fold cross-validation against in situ data and the best variation of model achieved RMSE, R2 and MAE of 0.0518 m3/m3, 0.7312 and 0.0374 m3/m3, respectively. The methodology proved to be useful for generating high-resolution SSM products, although additional improvements are necessary.

scholarly journals Evaluation of Remotely-Sensed and Model-Based Soil Moisture Products According to Different Soil Type, Vegetation Cover and Climate Regime Using Station-Based Observations over Turkey

2019 ◽  
Vol 11 (16) ◽  
pp. 1875 ◽  
Author(s):  
Burak Bulut ◽  
M. Tugrul Yilmaz ◽  
Mehdi H. Afshar ◽  
A. Ünal Şorman ◽  
İsmail Yücel ◽  
...  
Keyword(s):  
Soil Moisture ◽  
Root Mean Square ◽  
Soil Type ◽  
Vegetation Cover ◽  
Remotely Sensed ◽  
Mean Square ◽  
Climate Regime ◽  
Model Based ◽  
In Situ Observations

This study evaluates the performance of widely-used remotely sensed- and model-based soil moisture products, including: The Advanced Scatterometer (ASCAT), the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E), the European Space Agency Climate Change Initiative (ESA-CCI), the Antecedent Precipitation Index (API), and the Global Land Data Assimilation System (GLDAS-NOAH). Evaluations are performed between 2008 and 2011 against the calibrated station-based soil moisture observations collected by the General Directorate of Meteorology of Turkey. The calibration of soil moisture observing sensors with respect to the soil type, correction of the soil moisture for the soil temperature, and the quality control of the collected measurements are performed prior to the evaluation of the products. Evaluation of remotely sensed- and model-based soil moisture products is performed considering different characteristics of the time series (i.e., seasonality and anomaly components) and the study region (i.e., soil type, vegetation cover, soil wetness and climate regime). The systematic bias between soil moisture products and in situ measurements is eliminated by using a linear rescaling method. Correlations between the soil moisture products and the in situ observations vary between 0.57 and 0.87, while the root mean square errors of the products versus the in situ observations vary between 0.028 and 0.043 m3 m−3. Overall, according to the correlation and root mean square error values obtained in all evaluation categories, NOAH and ESA-CCI soil moisture products perform better than all the other model- and remotely sensed-based soil moisture products. These results are valid for the entire study time period and all of the sub-categories under soil type, vegetation cover, soil wetness and climate regime.

scholarly journals Multilayer Soil Moisture Mapping at a Regional Scale from Multisource Data via a Machine Learning Method

2019 ◽  
Vol 11 (3) ◽  
pp. 284 ◽  
Author(s):  
Linglin Zeng ◽  
Shun Hu ◽  
Daxiang Xiang ◽  
Xiang Zhang ◽  
Deren Li ◽  
...  
Keyword(s):  
Machine Learning ◽  
Soil Moisture ◽  
Regional Scale ◽  
Remotely Sensed ◽  
Temporal Variations ◽  
Training Data ◽  
Estimation Accuracy ◽  
Learning Approaches ◽  
Remotely Sensed Data ◽  
Deep Soil

Soil moisture mapping at a regional scale is commonplace since these data are required in many applications, such as hydrological and agricultural analyses. The use of remotely sensed data for the estimation of deep soil moisture at a regional scale has received far less emphasis. The objective of this study was to map the 500-m, 8-day average and daily soil moisture at different soil depths in Oklahoma from remotely sensed and ground-measured data using the random forest (RF) method, which is one of the machine-learning approaches. In order to investigate the estimation accuracy of the RF method at both a spatial and a temporal scale, two independent soil moisture estimation experiments were conducted using data from 2010 to 2014: a year-to-year experiment (with a root mean square error (RMSE) ranging from 0.038 to 0.050 m3/m3) and a station-to-station experiment (with an RMSE ranging from 0.044 to 0.057 m3/m3). Then, the data requirements, importance factors, and spatial and temporal variations in estimation accuracy were discussed based on the results using the training data selected by iterated random sampling. The highly accurate estimations of both the surface and the deep soil moisture for the study area reveal the potential of RF methods when mapping soil moisture at a regional scale, especially when considering the high heterogeneity of land-cover types and topography in the study area.

scholarly journals Global soil moisture data derived through machine learning trained with in-situ measurements

2021 ◽  
Vol 8 (1) ◽  
Author(s):  
Sungmin O. ◽  
Rene Orth
Keyword(s):  
Machine Learning ◽  
Soil Moisture ◽  
Large Scale ◽  
Short Term Memory ◽  
Temporal Dynamics ◽  
Soil Moisture Data ◽  
Wide Range ◽  
Global Soil

AbstractWhile soil moisture information is essential for a wide range of hydrologic and climate applications, spatially-continuous soil moisture data is only available from satellite observations or model simulations. Here we present a global, long-term dataset of soil moisture derived through machine learning trained with in-situ measurements, SoMo.ml. We train a Long Short-Term Memory (LSTM) model to extrapolate daily soil moisture dynamics in space and in time, based on in-situ data collected from more than 1,000 stations across the globe. SoMo.ml provides multi-layer soil moisture data (0–10 cm, 10–30 cm, and 30–50 cm) at 0.25° spatial and daily temporal resolution over the period 2000–2019. The performance of the resulting dataset is evaluated through cross validation and inter-comparison with existing soil moisture datasets. SoMo.ml performs especially well in terms of temporal dynamics, making it particularly useful for applications requiring time-varying soil moisture, such as anomaly detection and memory analyses. SoMo.ml complements the existing suite of modelled and satellite-based datasets given its distinct derivation, to support large-scale hydrological, meteorological, and ecological analyses.

scholarly journals Global downscaling of remotely sensed soil moisture using neural networks

2018 ◽  
Vol 22 (10) ◽  
pp. 5341-5356 ◽  
Author(s):  
Seyed Hamed Alemohammad ◽  
Jana Kolassa ◽  
Catherine Prigent ◽  
Filipe Aires ◽  
Pierre Gentine
Keyword(s):  
Soil Moisture ◽  
Spatial Resolution ◽  
Land Surface ◽  
Vegetation Index ◽  
Spatial Scales ◽  
Ancillary Data ◽  
Repeat Time ◽  
Spatiotemporal Scales ◽  
Better Than

Abstract. Characterizing soil moisture at spatiotemporal scales relevant to land surface processes (i.e., of the order of 1 km) is necessary in order to quantify its role in regional feedbacks between the land surface and the atmospheric boundary layer. Moreover, several applications such as agricultural management can benefit from soil moisture information at fine spatial scales. Soil moisture estimates from current satellite missions have a reasonably good temporal revisit over the globe (2–3-day repeat time); however, their finest spatial resolution is 9 km. NASA's Soil Moisture Active Passive (SMAP) satellite has estimated soil moisture at two different spatial scales of 36 and 9 km since April 2015. In this study, we develop a neural-network-based downscaling algorithm using SMAP observations and disaggregate soil moisture to 2.25 km spatial resolution. Our approach uses the mean monthly Normalized Differenced Vegetation Index (NDVI) as ancillary data to quantify the subpixel heterogeneity of soil moisture. Evaluation of the downscaled soil moisture estimates against in situ observations shows that their accuracy is better than or equal to the SMAP 9 km soil moisture estimates.

scholarly journals Quality Improvement of Satellite Soil Moisture Products by Fusing with In-Situ Measurements and GNSS-R Estimates in the Western Continental U.S.

2018 ◽  
Vol 10 (9) ◽  
pp. 1351 ◽  
Author(s):  
Hongzhang Xu ◽  
Qiangqiang Yuan ◽  
Tongwen Li ◽  
Huanfeng Shen ◽  
Liangpei Zhang ◽  
...  
Keyword(s):  
Soil Moisture ◽  
Water Cycle ◽  
Satellite System ◽  
Global Scale ◽  
Generalized Regression Neural Network ◽  
Level 3 ◽  
Global Navigation Satellite ◽  
Microwave Sensors ◽  
Surface Fusion

Soil moisture is a key component of the water cycle budget. Sensing soil moisture using microwave sensors onboard satellites is an effective way to retrieve surface soil moisture (SSM) at a global scale, but the retrieval accuracy in some regions is inadequate due to the complicated factors influencing the general retrieval process. On the other hand, monitoring soil moisture directly through in-situ devices is capable of providing high-accuracy SSM measurements, but the distribution of such stations is sparse. Recently, the Global Navigation Satellite System interferometric Reflectometry (GNSS-R) method was used to derive field-scale SSM, which can serve as a supplement to contemporary sparse in-situ soil moisture networks. On this basis, it is of great research significance to explore the fusion of these different kinds of SSM data, so as to improve the present satellite SSM products with regard to their data accuracy. In this paper, a multi-source point-surface fusion method based on the generalized regression neural network (GRNN) model is applied to fuse the Soil Moisture Active Passive (SMAP) Level 3 radiometer SSM daily product with in-situ measured and GNSS-R estimated SSM data from five soil moisture networks in the western continental U.S. The results show that the GRNN model obtains a fairly good performance, with a cross-validation R value of approximately 0.9 and a ubRMSE of 0.044 cm3 cm−3. Furthermore, the fused SSM product agrees well with the site-specific SSM data in terms of time and space, which demonstrates that the proposed GRNN model is able to construct the non-linear relationship between the point- and surface-scale SSM.

scholarly journals An Approach for Downscaling SMAP Soil Moisture by Combining Sentinel-1 SAR and MODIS Data

2019 ◽  
Vol 11 (23) ◽  
pp. 2736 ◽  
Author(s):  
Jueying Bai ◽  
Qian Cui ◽  
Wen Zhang ◽  
Lingkui Meng
Keyword(s):  
Soil Moisture ◽  
Vegetation Cover ◽  
Land Surface ◽  
Feature Space ◽  
Entire Study ◽  
Vegetation Indexes ◽  
Input Variables ◽  
Entire Study Area ◽  
Infrared Data

A method is proposed for the production of downscaled soil moisture active passive (SMAP) soil moisture (SM) data by combining optical/infrared data with synthetic aperture radar (SAR) data based on the random forest (RF) model. The method leverages the sensitivity of active microwaves to surface SM and the triangle/trapezium feature space among vegetation indexes (VIs), land surface temperature (LST), and SM. First, five RF architectures (RF1–RF5) were trained and tested at 9 km. Second, a comparison was performed for RF1–RF5, and were evaluated against in situ SM measurements. Third, two SMAP-Sentinel active–passive SM products were compared at 3 km and 1 km using in situ SM measurements. Fourth, the RF5 model simulations were compared with the SMAP L2_SM_SP product based on the optional algorithm at 3 km and 1 km resolutions. The results showed that the downscaled SM based on the synergistic use of optical/infrared data and the backscatter at vertical–vertical (VV) polarization was feasible in semi-arid areas with relatively low vegetation cover. The RF5 model with backscatter and more parameters from optical/infrared data performed best among the five RF models and was satisfactory at both 3 km and 1 km. Compared with L2_SM_SP, RF5 was more superior at 1 km. The input variables in decreasing order of importance were backscatter, LST, VIs, and topographic factors over the entire study area. The low vegetation cover conditions probably amplified the importance of the backscatter and LST. A sufficient number of VIs can enhance the adaptability of RF models to different vegetation conditions.

scholarly journals Assessment of Multi-Scale SMOS and SMAP Soil Moisture Products across the Iberian Peninsula

2020 ◽  
Vol 12 (3) ◽  
pp. 570 ◽  
Author(s):  
Gerard Portal ◽  
Thomas Jagdhuber ◽  
Mercè Vall-llossera ◽  
Adriano Camps ◽  
Miriam Pablos ◽  
...  
Keyword(s):  
Soil Moisture ◽  
Iberian Peninsula ◽  
Spatial Scale ◽  
Temporal Dynamics ◽  
Microwave Radiometer ◽  
Added Value ◽  
Optical Data ◽  
Enhanced Resolution ◽  
The Impact

In the last decade, technological advances led to the launch of two satellite missions dedicated to measure the Earth’s surface soil moisture (SSM): the ESA’s Soil Moisture and Ocean Salinity (SMOS) launched in 2009, and the NASA’s Soil Moisture Active Passive (SMAP) launched in 2015. The two satellites have an L-band microwave radiometer on-board to measure the Earth’s surface emission. These measurements (brightness temperatures TB) are then used to generate global maps of SSM every three days with a spatial resolution of about 30–40 km and a target accuracy of 0.04 m3/m3. To meet local applications needs, different approaches have been proposed to spatially disaggregate SMOS and SMAP TB or their SSM products. They rely on synergies between multi-sensor observations and are built upon different physical assumptions. In this study, temporal and spatial characteristics of six operational SSM products derived from SMOS and SMAP are assessed in order to diagnose their distinct features, and the rationale behind them. The study is focused on the Iberian Peninsula and covers the period from April 2015 to December 2017. A temporal inter-comparison analysis is carried out using in situ SSM data from the Soil Moisture Measurements Station Network of the University of Salamanca (REMEDHUS) to evaluate the impact of the spatial scale of the different products (1, 3, 9, 25, and 36 km), and their correspondence in terms of temporal dynamics. A spatial analysis is conducted for the whole Iberian Peninsula with emphasis on the added-value that the enhanced resolution products provide based on the microwave-optical (SMOS/ERA5/MODIS) or the active–passive microwave (SMAP/Sentinel-1) sensor fusion. Our results show overall agreement among time series of the products regardless their spatial scale when compared to in situ measurements. Still, higher spatial resolutions would be needed to capture local features such as small irrigated areas that are not dominant at the 1-km pixel scale. The degree to which spatial features are resolved by the enhanced resolution products depend on the multi-sensor synergies employed (at TB or soil moisture level), and on the nature of the fine-scale information used. The largest disparities between these products occur in forested areas, which may be related to the reduced sensitivity of high-resolution active microwave and optical data to soil properties under dense vegetation.

Exploring the use of machine-learning techniques to integrate ground- and remote sensing-based observations for efficient near-surface soil moisture mapping

2020 ◽  
Author(s):  
Sarah Schönbrodt-Stitt ◽  
Paolo Nasta ◽  
Nima Ahmadian ◽  
Markus Kurtenbach ◽  
Christopher Conrad ◽  
...  
Keyword(s):  
Machine Learning ◽  
Soil Moisture ◽  
Spatial Resolution ◽  
Surface Soil ◽  
Bare Soil ◽  
Small Scale ◽  
Surface Soil Moisture ◽  
Near Surface ◽  
Campania Region

<p>Mapping near-surface soil moisture (<em>θ</em>) is of tremendous relevance for a broad range of environment-related disciplines and meteorological, ecological, hydrological and agricultural applications. Globally available products offer the opportunity to address <em>θ</em> in large-scale modelling with coarse spatial resolution such as at the landscape level. However, <em>θ</em> estimation at higher spatial resolution is of vital importance for many small-scale applications. Therefore, we focus our study on a small-scale catchment (MFC2) belonging to the “Alento” hydrological observatory, located in southern Italy (Campania Region). The goal of this study is to develop new machine-learning approaches to estimate high grid-resolution (about 17 m cell size) <em>θ</em> maps from mainly backscatter measurements retrieved from C-band Synthetic Aperture Radar (SAR) based on Sentinel-1 (S1) images and from gridded terrain attributes. Thus, a workflow comprising a total of 48 SAR-based <em>θ</em> patterns estimated for 24 satellite overpass dates (revisit time of 6 days) each with ascendant and descendent orbits will be presented. To enable for the mapping, SAR-based <em>θ</em> data was calibrated with in-situ measurements carried out with a portable device during eight measurement campaigns at time of satellite overpasses (four overpass days in total with each ascendant and descendent satellite overpasses per day in November 2018). After the calibration procedure, data validation was executed from November 10, 2018 till March 28, 2019 by using two stationary sensors monitoring <em>θ</em> at high-temporal (1-min recording time). The specific sensor locations reflected two contrasting field conditions, one bare soil plot (frequently kept clear, without disturbance of vegetation cover) and one non-bare soil plot (real-world condition). Point-scale ground observations of <em>θ</em> were compared to pixel-scale (17 m × 17 m), SAR-based <em>θ</em> estimated for those pixels corresponding to the specific positions of the stationary sensors. Mapping performance was estimated through the root mean squared error (RMSE). For a short-term time series of <em>θ</em> (Nov 2018) integrating 136 in situ, sensor-based <em>θ</em> (<em>θ</em><sub>insitu</sub>) and 74 gravimetric-based <em>θ</em> (<em>θ</em><sub>gravimetric</sub>) measurements during a total of eight S1 overpasses, mapping performance already proved to be satisfactory with RMSE=0.039 m³m<sup>-</sup>³ and R²=0.92, respectively with RMSE=0.041 m³m<sup>-</sup>³ and R²=0.91. First results further reveal that estimated satellite-based <em>θ</em> patterns respond to the evolution of rainfall. With our workflow developed and results, we intend to contribute to improved environmental risk assessment by assimilating the results into hydrological models (e.g., HydroGeoSphere), and to support future studies on combined ground-based and SAR-based <em>θ</em> retrieval for forested land (future missions operating at larger wavelengths e.g. NISARL-band, Biomass P-band sensors).</p>

Sign in / Sign up

Export Citation Format

Share Document