Multi-sensor SAR Data Research Articles

Urbanization and land subsidence: multi-decadal investigation combined SBAS-InSAR and multi-factors in Shanghai, China

Situated in the Yangtze River Delta’s alluvial plain, Shanghai has faced intensified land subsidence problems since the economic reforms of the 1990s. However, previous studies have lacked long time-series land subsidence monitoring and an analysis of multifactorial interactions in Shanghai over the past years. This study analyzed Shanghai’s land subsidence from 1992 to 2020 using SAR data from ERS-1/2, ENVISAT ASAR, and Sentinel-1A, applying the SBAS-InSAR technique. The geo-detector method was used to identify key factors affecting land subsidence, including groundwater, urban structure, metro development, and population dynamics. The results showed that (1) the utilization of multi-sensor SAR data enables the comprehensive analysis of urban land subsidence over an extended temporal series, facilitating the acquisition of its spatiotemporal distribution; (2) the extent of subsidence in Shanghai was gradually spreading from the core to the outskirts, and the rate of subsidence was gradually decreasing during 1992-2020 period; (3) groundwater was the most influential factor on land subsidence, and the interaction between groundwater and metro had a stronger influence on land subsidence.

Gangotri glacier dynamics from multi-sensor SAR and optical data

The present study has analyzed dynamics of Gangotri glacier using multiple remote sensing (RS) datasets and ground based observations. Interferometric Synthetic Aperture Radar (InSAR) data pairs from European Remote Sensing satellite (ERS 1/2) tandem pair for spring of 1996, Sentinel-1 SAR pairs and Japanese’s Advance Land Observation System (ALOS) PALSAR-2 SAR data for Spring of 2015 were used to derive glacier-surface velocity at seasonal time scale using Differential InSAR (DInSAR) techniques. Bi-static TanDEM-X (Experimental) data was used for the 1st time to estimate glacier surface elevation changes for a period of 22, 44, 88 days during summer of 2012 using InSAR techniques in this study. Annual glacier velocity was also estimated using temporal panchromatic data of LANDSAT-5 (30 m), LANDSAT-7/8 (15 m), Sentinel-2 (10 m) and Indian Remote Sensing Satellite IRS-1C/1D panchromatic (5 m) data during 1998–2019 with feature tracking approach. This study has estimated glacier surface velocity and surface elevation changes for the major parts of Gangotri glacier and its tributary glaciers using medium to high resolution optical and SAR datasets, at annual and seasonal time scale, which is an improvement over earlier studies, wherein snout based glacier recession or only main glacier velocities were reported. The velocity and slope were used to assess glacier-ice thickness distribution using Glabtop-2, slope dependent and laminar flow based methods over the Gangotri group of glaciers. The estimated ice thickness was estimated in the range of 58–550 m for the complete glacier while few small areas in middle & upper regions carry higher thickness of about 607 m. The estimated glacier-ice thickness was found in the range of 58–67 m at the snout region. The estimation was validated using 2014 field measurements from Terrestrial Laser Scanner (TLS) for the first time and correlation was found to be 0.799 at snout of the glacier.

Long-term spatiotemporal evolution of land subsidence in Konya metropolitan area (Turkey) based on multisensor SAR data

We have studied the spatiotemporal evolution of surface deformation in Konya city and its vicinity using advanced multitemporal synthetic aperture radar techniques with SAR data acquired by Envisat, ALOS-1, and Sentinel-1 A/B satellites between 2004 and 2020. Velocity maps and time series show that the city has been subsiding with varying rates in space and time since 2004. The pattern of deformation shows two main lobes of subsidence centered in the western and eastern sides of the city with a nondeforming north-south trending narrow zone in between. Subsidence rate increases from a few cm/yr to 11 cm/yr between 2014 and 2019. As of 2019, subsidence has slowed down dramatically, giving rise to uplift in some places. Spatiotemporal variation of subsidence and its strong correlation with change in water table level confirm the inferences that subsidence in the metropolitan area of Konya is due to over drafting of the ground water for urban needs. The decrease in subsidence rate over the last two years appears to be due to the city?s residents supplying their water from recently built dams instead of aquifers beneath the city. Initial excessive groundwater extraction in agricultural areas caused ~4 m drops every year in the water table level, which, in turn, gave rise to 8 cm subsidence every year. Modeling of the subsidence shows 7.7 x 106 m3/yr volume loss due to compaction of the aquifer in the Konya metropolitan area and its vicinity between 2014 and 2018.

Evaluation of different machine learning algorithms for pearl millet discrimination using multi-sensor SAR data

This study aims to assess the potential of multi-temporal Sentinel-1 and RADARSAT-2 data for pearl millet discrimination using various classifiers—machine learning and knowledge-based decision tree (m-DT and k-DT), Support Vector Machine (SVM), and Random Forest (RF). Results show that the RF classifier outperformed the other classifiers in terms of overall accuracy (OA) and kappa coefficient. The highest Overall Accuracy (OA) of 89.1% and kappa coefficients of 0.822 and 0.814 were achieved by the RF classifier for both the Sentinel-1 and RADARSAT-2 data, respectively. The Variable Importance (VI) of the RF classifier revealed that the polarimetric parameters—volume scattering, double bounce, entropy and alpha angle were crucial for pearl millet discrimination. Polarimetric parameters, when ingested with the RF machine learning classifier, achieved better classification accuracies.

Least Square Based Iteration Approach for Agricultural Soil Moisture Retrieval Using Multi-Sensor Data

The main objective of this study is to develop a robust soil moisture retrieval approach using multi-sensor remote sensing data. Firstly, the water cloud model was employed to eliminate the vegetation effects on SAR observations over vegetated areas, thus to obtain the bare soil backscatter associated with soil moisture. Then, against the underdetermined system for soil moisture retrieval, the advanced integral equation model and calibrated integral equation model were integrated to construct soil moisture retrieval scheme in combination with multi-sensor SAR observations. Through the above processing the influences of vegetation and surface roughness can be minimized in the developed soil moisture retrieval scheme. Finally, the least square based iteration approach was applied to derive soil moisture with multi-sensor SAR observations as inputs. Based on the field measurements and multi-sensor SAR data (C-band and X-band). acquired on June 2015 and October 2015 over Hebei agricultural areas, quantitative evaluation of the developed approach was implemented. The results indicate that accurate soil moisture was obtained by the developed approach over corn covered area and bare agricultural area. In comparison with the results obtained by LUT method, the least square based iteration approach achieved better estimations for soil moisture.

Estimation of crop angle of inclination for lodged wheat using multi-sensor SAR data

Lodging - the bending of crop stems - reduces the quantity and quality of cereal crop yields. Early quantification of crop lodging is important to prevent further losses and to facilitate harvesting operations. Crop angle of inclination (CAI) is a quantitative measure of the lodging stage and a component of lodging severity/score. CAI is an important structural parameter for lodged crops and very few studies have investigated its estimation using satellite-based remote sensing. In this study, the performance of Sentinel-1 and multi-incidence angle (FQ8-27° and FQ21-41°) RADARSAT-2 data were investigated for estimating CAI. Temporal crop biophysical/structural parameters (CAI and crop height) and meteorological data (rainfall and wind speed) were collected throughout May 1-June 30, 2018 in a very large commercial farm located in Jolanda di Savoia, Ferrara, Italy. Field data were grouped into different crop lodging stages (non-lodged/healthy (H), moderate lodging (ML), severe lodging (SL) and very severe lodging (VSL)) based on CAI. Quantitative relationships were established between field-measured CAI values and the RS-derived metrics for Sentinel-1 and RADARSAT-2 timeseries using support vector regression (SVR) models. The RADARSAT-2 FQ8 model performed most robustly with a R2CV (cross-validated R2) of 0.87 and a RMSECV (cross-validated RMSE) of 8.89° while the performance of the Sentinel-1 and RADARSAT-2 FQ21 models were comparable with an RMSECV of 11.35° and 11.63° respectively. Low incidence angle data were particularly sensitive to high CAI values (VSL) while high incidence angle data were useful for predicting lower CAI (ML and SL). While the RADARSAT-2 FQ-8 model outperformed the other two, the Sentinel-1 model still explained 78% of the CAI variability in the study site, which is important in the context of operational crop lodging stage assessment. This is the first study to demonstrate the utility of SAR remote sensing data for estimating CAI as a measure of the lodging stage and a component of lodging severity.

SAR-SHARPENING IN THE KENNAUGH FRAMEWORK APPLIED TO THE FUSION OF MULTI-MODAL SAR AND OPTICAL IMAGES

Abstract. The Kennaugh framework turned out to be a powerful tool for the preparation of multi-sensor SAR data during the last years. Using intensity-based (an-) isotropic diffusion algorithms like the Multi-scale Multi-looking or the Schmittlets, even robust pre-classification change detection from multi-polarized images is enabled. The only missing point so far, namely the integration of multi-mode SAR data in one image, is accomplished in this article. Furthermore, the Kennaugh decomposition is extended to multi-spectral data as well. Hence, arbitrary Kennaugh elements, be it from SAR or optical images, can be fused. The mathematical description of the most general image fusion is derived and applied to four scenarios. The validation section considers the distribution of mean and gradient in the original and the fused images by the help of scatter plots. The results prove that the fused images adopt the spatial gradient of the input image with a higher geometric resolution and preserve the local mean of the input image with a higher polarimetric and thus also radiometric resolution. Regarding the distribution of the entropy and alpha angle, the fused images are always characterized by a higher variance in the entropy-alpha-plane and therewith, a higher resolution in the polarimetric domain. The proposed algorithm guarantees optimal information integration while ensuring the separation of intensity and polarimetric/spectral information. The Kennaugh framework is ready now to be used for the sharpening of multi-sensor image data in the spatial, radiometric, polarimetric, and even spectral domain.

Advanced SBAS-DInSAR Technique for Controlling Large Civil Infrastructures: An Application to the Genzano di Lucania Dam.

Monitoring surface deformation on dams is commonly carried out by in situ geodetic surveying, which is time consuming and characterized by some limitations in space coverage and frequency. More recently microwave satellite-based technologies, such as advanced-DInSAR (Differential Synthetic Aperture Radar Interferometry), have allowed the integration and improvement of the observation capabilities of ground-based methods thanks to their effectiveness in collecting displacement measurements on many non-destructive control points, corresponding to radar reflecting targets. The availability of such a large number of points of measurement, which are distributed along the whole structure and are characterized by millimetric accuracy on displacement rates, can be profitably adopted for the calibration of numerical models. These models are implemented to simulate the structural behaviour of a dam under conditions of stress thus improving the ability to maintain safety standards. In this work, after having analysed how advanced DInSAR can effectively enhance the results from traditional monitoring systems that provide comparable accuracy measurements on a limited number of points, an FEM model of the Genzano di Lucania earth dam is developed and calibrated. This work is concentrated on the advanced DInSAR technique referred to as Small BAseline Subset (SBAS) approach, benefiting from its capability to generate deformation time series at full spatial resolution and from multi-sensor SAR data, to measure the vertical consolidation displacement of the Genzano di Lucania earth dam.

GROUND SUBSIDENCE MONITORING WITH MT-InSAR AND MECHANISM INVERSION OVER XI’AN, CHINA

Abstract. The ancient Xi’an, China, has been suffering severe land subsidence and ground fissure hazards since the 1960s, which has affected the safety of Subways. Multi-sensor SAR data are conducted to monitor the latest complex ground deformation and its influence on subway line No.3 over Xi’an. Annual deformation rates have been retrieved to reveal the spatiotemporal evolution of ground subsidence in Xi’an city from 2013 to 2017. Meanwhile, the correlation between land subsidence and ground fissures are analyzed by retrieving the deformation differences in both sides of the fissures. Besides, the deformation along subway line No. 3 is analyzed, and the fast deformation section is quantitatively studied. Finally, a flat lying sill model with distributed contractions is implemented to model the InSAR deformation over YHZ subsidence center, which manifests that the ground deformation is mainly caused by groundwater withdrawal.

Multitemporal Soil Moisture Retrieval over Bare Agricultural Areas by Means of Alpha Model with Multisensor SAR Data

The objective of this research is to optimize the Alpha approximation model for soil moisture retrieval using multitemporal SAR data. The Alpha model requires prior knowledge of soil moisture range to constrain soil moisture estimation. The solution of the Alpha model is an undetermined problem due to the fact that the number of observation equations is less than the number of unknown parameters. This research primarily focused on the optimization of Alpha model by employing multisensor and multitemporal SAR data. The disadvantage of the Alpha model can be eliminated by the combination of multisensor SAR data. The optimized Alpha model was evaluated on the basis of a comprehensive campaign for soil moisture retrieval, which acquired multisensor time series SAR data and coincident field measurements. The agreement between the estimated and measured soil moisture was within a root mean square error of 0.08 cm3/cm3 for both methods. The optimized Alpha model shows an obvious improvement for soil moisture retrieval. The results demonstrated that multisensor and multitemporal SAR data are favorable for time series soil moisture retrieval over bare agricultural areas.

Enhanced landslide investigations through advanced DInSAR techniques: The Ivancich case study, Assisi, Italy

We extensively exploit advanced Differential SAR Interferometry (DInSAR) techniques for enhanced landslide investigations. We focus on the Ivancich area, Assisi, Central Italy, which is affected by a deep-seated landslide investigated through in-situ surveys. For this area, large data sets of SAR acquisitions were collected by the C-band ERS-1/2 and ENVISAT sensors (from April 1992 to November 2010), and by the X-band radars of the COSMO-SkyMed (CSK) constellation (from December 2009 to February 2012). We concentrate on the advanced DInSAR technique referred to as Small BAseline Subset (SBAS) approach, benefiting of its capability to generate deformation time series at full spatial resolution and from multi-sensor SAR data. This allows us to present one of the first examples for a landslide area of ERS-1/2 — ENVISAT deformation time series exceeding 18 years. The results allowed characterizing the long-term behaviour of the landslide, and identifying sectors of the unstable slope affected by different deformation dynamics. Analysis of the CSK data set, characterized by a reduced revisit time and improved spatial resolution, resulted in a 15-time larger point density with respect to the ERS-ENVISAT measurements, allowing to investigate nearly all the buildings (and, in many cases, portions of buildings) in the landslide area. Lastly, we present an innovative modelling approach based on the effective integration of the DInSAR measurements with traditional geological and geotechnical information, providing deeper insights on the kinematical evolution of the landslide. We consider our analysis a prototype example that can be extended to different geological and geotechnical conditions, providing significant advances in the understanding of ground deformations induced by active landslides.

Multi-sensor advanced DInSAR monitoring of very slow landslides: The Tena Valley case study (Central Spanish Pyrenees)

In this work it is demonstrated how the combination of multi-sensor and multi-temporal SAR data improves the monitoring capacity of very slow landslides. The study area is the Upper Tena Valley (Central Spanish Pyrenees), where mostly very slow earth flows, have caused direct damage estimated at approximately 15M€ in the last decade. The advanced DInSAR processing of ascending orbit ALOS PALSAR images (2006–2010) and descending orbit ERS & Envisat (2001–2007) and TerraSAR-X (2008) datasets, has provided heterogeneous displacement velocity measurements. The projection along the steepest slope of LOS displacements reduces the geometrical differences introduced by each satellite. These results were compared and validated with available D-GPS measurements. In a second step PS displacement data are combined with a landslide inventory, increasing the number of monitored landslides with sufficient PS from 4% to 19%, obtained from C- and X-band data, to 38% of the total (294). Finally, the retrieved multi-sensor landslides velocities are classified with respect to the magnitude of road damages that occurred in the 2008–2010 period. At local scale, the combination of multi-sensor data was useful to differentiate different landslide displacement directions, to measure different velocity patterns within the same moving mass, and to distinguish the slower (natural) and faster (human induced) landslides. The analysis of displacement time series permitted to evaluate the acceleration caused by a destabilising anthropogenic change, and the correlation between displacement variability and seasonal precipitation and melting.