Bandung is a capital city of West Java Province, one of the largest city by population in Indonesia. The city experience rapid development of industrialization since 1970s. The growth of population increase afterwards, and create demand of resources to support the city, including fresh water. Ground water has become inevitable source to fill the high demand while on the other hand, government cannot meet the demand of fresh water. This supply gap causes exploitation of ground water, and triggered environment quality degradation, which can be indicated by high land subsidence rate. We processed ALOS-PALSAR 1 data in the period of 2006-2011, then Sentinel-1 for period of 2014-2020. Small Baseline Subset Analysis (SBAS) is used for time-series analysis. GPS annual campaign were also done for period of 2005-2018, and we also manage to put several low-cost GPS stations in some prominent points to calibrate InSAR data processing. Our results shows that between 2006-2011, the area of land subsidence mostly located on the industrial area, with the highest rate up to 13 cm/year in Leuwi Gajah Industrial Complex. Some other areas which suffer land subsidence during the same period, are also located in the industrial complex such as Gede Bage, Ranca Ekek, and Majalaya. However, the spatial distribution of land subsidence is shifting during the period of 2011-2015. The demand of ground water from household grows rapidly within this period, indicated by a much higher rate of land subsidence occue in housing area. The area and rate of land subsidence broaden significantly in some housing areas such as Kopo and Margahayu with subsidence rate of 12 cm/year and 8 cm/year respectively. The opposite trend happened in the previously stated area, slowing down of land subsidence rate are observed in several locations with 5-8 cm/year slower compare to 2006-2011 period. The impact of land subsidence in housing area has caused the widen of flooding area and degradation of ground water quality, especially in Kopo District which suffers highest subsidence rate in housing area.
Many surface and deep aquifers in Central Mexico are overexploited to address water needs for public/private and industrial use, with boosting demand in expanding cities and metropolises. As a consequence, aquifers deplete and urban centers are widely affected by land subsidence and its derived risk for housing, transport and other infrastructure, with associated economic loss, thus representing a key topic of concern for inhabitants, authorities and stakeholders. This work analyses groundwater resource availability and aquifer storage change in Central Mexico based on groundwater management reports, issued by the National Water Commission and published yearly in Mexico's Official Federal Gazette. Evidence from piezometric measurements and aquifer modeling are discussed in relation with satellite InSAR surveys based on geospatial analysis of Sentinel-1 IW SAR big data processed within ESA’s Geohazards Exploitation Platform (GEP), using the on-demand Parallel Small BAseline Subset (P-SBAS) service. Case studies include a number of aquifers where significant groundwater deficits and aquifer storage changes were estimated over the last years, including those in the Metropolitan Area of Mexico City [1], one of the fastest sinking cities globally (up to 40 cm/year subsidence rates); the state of Aguascalientes [2], where a structurally-controlled fast subsidence process (over 10 cm/year rates) affects the namesake valley and capital city; and the Metropolitan Area of Morelia [3], a rapidly expanding metropolis where population doubled over the last 30 years and a subsidence-creep-fault process has been identified. Surface faulting hazard resulting from differential settlement is constrained via the estimation of angular distortions that are produced on urban structures using the InSAR-derived vertical deformation field, as well as the computation of horizontal strain (tensile and compressive) based on the InSAR-derived E-W deformation field. A methodology to embed such information into the process of risk assessment for urban infrastructure is proposed and demonstrated. The results and discussion will showcase how the InSAR deformation datasets and their derived products are essential not only to constrain the deformation processes, but also to provide a valuable input for the quantification of properties and population at risk. InSAR-derived evidence of land subsidence and its induced risk could be a crucial information source for groundwater management policy makers and regulators, to identify the most impacted cities and optimize groundwater management and development plans to accommodate existing and future water demands.
[1] Cigna F., Tapete D. 2021. Present-day land subsidence rates, surface faulting hazard and risk in Mexico City with 2014-2020 Sentinel-1 IW InSAR. Remote Sensing of Environment, 253, 1-19, https://doi.org/10.1016/j.rse.2020.112161
[2] Cigna F., Tapete D. 2021. Satellite InSAR survey of structurally-controlled land subsidence due to groundwater exploitation in the Aguascalientes Valley, Mexico. Remote Sensing of Environment, 254, 1-23, https://doi.org/10.1016/j.rse.2020.112254
[3] Cigna F., Osmanoǧlu B., Cabral-Cano E., Dixon T.H., Ávila-Olivera J.A., Garduño-Monroy V.H., DeMets C., Wdowinski S., 2012. Monitoring land subsidence and its induced geological hazard with Synthetic Aperture Radar Interferometry: A case study in Morelia, Mexico. Remote Sensing of Environment, 117, 146-161. https://doi.org/10.1016/j.rse.2011.09.005
United Arab Emirates (UAE) characterised by arid climate with limited resources of fresh water and high-water demand in sectors of domestic, agriculture, and industry. Due to this limitation in water resources, sustainable groundwater practices are required to maintain the available resources to diminish. One of the most crucial groundwater practices are monitoring of groundwater dynamics, in quality and quantity, and the implications of unsustainable groundwater usage.
Al Ain region is located at the eastern part of Abu Dhabi Emirate, UAE at the border with Sultanate Oman. This region occupies 50% of the Abu Dhabi Emirate agricultural activities consumes huge amount of groundwater with annual discharge more than 200 million m3. The groundwater resources can be found at the unconfined gravel aquifers and sand dune aquifers.
The current study aims at investigating the deformations occurring at the site due to the overexploitation of the aquifers by combining SAR satellite data, with ground water level measurements and ground truth surveys. Sentinel-1 data, provided by the European Space Agency (ESA), were used to process the SAR interferometry over the study area. Water level data were provided by the Environment Agency of Abu Dhabi (EAD), and they were used to determine zones affected by groundwater overexploitation. Land surfaced subsidence evidences were identified in the field, confirming the deformations identified by the SAR interferometry product. The dataset used consists of 37 Sentinel-1A Single Look Complex (SLC) images acquired along the ascending orbit from path 130 and frames 73 and 75, for a time span between February 2015 and May 2019. The image acquired on 22 October 2017 was selected as a primary, or master, image to increase the expected coherence due to it is minimum spatial and temporal baselines.
The water level dataset indicated an extensive cone of depression covering the area under investigation from 2013 to 2019. It is clear that the extended network of irrigation wells has systematically affected the unconfined sand dune aquifers unit and resulted in lowering the groundwater level with a maximum drawdown at its centre of approximately 40 to 50 m. As expected, at the perimeter of the cone, the ground water lowering gradually decreases in relation to the distance from the centre of the cone. The great discharge from the aquifer, more than 240 million m3, along with a very low hydraulic conductivity of the aquifer resulted in a low annual groundwater recharge.
The study revealed an extensive land surface subsidence with a rate of 40 mm/year in the period between 2015 and 2019. The cone of depression for the water level drawdown in the study area was found in spatial correlation with the detected land surface subsidence bowl. This can be concluded that the land surface subsidence was triggered by the groundwater over extraction.
Furthermore, it was proved that the repeat-pass satellite SAR interferometry can provided substantial information about the actual extent of the land subsidence phenomenon. Space-based technologies are cost effective, providing high spatial coverage. So, they are able to fill the data and knowledge gaps and reduce the uncertainties by providing high spatial and temporal valuable information about the extend and the progress of the subsidence.
This work was supported by a grant from the United Arab Emirates University (UAEU) National Center for Water and Energy under grant number 31R155-Research Center-NWC-3-2017.
Background
The Netherlands has been actively managing its groundwater table for centuries, and much of the Dutch agricultural sector is based in drained peatlands called polders. The polders are separated into parcels which are rectangular plots surrounded by drainage ditches. It has been observed that groundwater levels are the most significant driver of soil surface height variation in the region [1]. It has been hypothesized that groundwater management regimes in these regions are causing them to irreversibly subside at a rate which is faster than sea level rise [2, 3].
Problem Statement
Direct observation of surface motion of this region has so far not been possible with distributed scatterer (DS) InSAR due to high levels of noise in these grass-covered fields and rapid deformation between consecutive SAR acquisitions [4]. These rapid shifts can frequently result in phase unwrapping errors when typical DS processing techniques are used. Additionally, relating the observed InSAR time series with respect to an absolute reference frame has not been possible due to the decoupling of point scatterer (PS) and DS processing techniques, and the lack of a well-defined reference benchmark. To understand the scope of the effects of subsidence in these regions, and subsequently act on it, scientists and policymakers need to know the subsidence rates in absolute terms, so that they may be compared to other locations, or other hazards such as sea level rise.
Our Approach
We adopt a novel multilooking strategy which uses parcels as the basic unit of measure for a DS InSAR monitoring system. This is a natural choice, as the land cover and water table are almost always consistent over a given parcel. Our approach has several advantages over traditional multilooking strategies, as it simultaneously groups pixels together which are physically acted upon by the same mechanism, while also providing a many-fold increase in the effective number of looks. This significantly reduces measurement noise and ensures that
all the pixels being multilooked are tied to the same ergodic process, which a typical boxcar filter will not do, even when coupled with a statistical homogeneity test.
These multilooked phases can now be treated as virtual point scatterers; they are assigned a location corresponding to the centroid of the parcel in question, and imported into the Delft Persistent Scatterer Interferometry (DePSI) system for time-series analysis as 2nd order points [5]. This allows us to take advantage of the robust point scatterers (PS) in the image to form a 1st order network of points for atmospheric phase screen (APS) removal, trend removal and variance component estimation. This mixed system allows for the simultaneous monitoring of both agricultural zones and the built environment.
Via the connection to the 1st order network, we are able to make an arc connection to the Integrated Geodetic Reference Station (IGRS) [6] in the region. These stations are comprised of a corner reflector, a GPS receiver and other geodetic instrumentation which will allow us to make a direct connection from the observed parcel movement to an absolute geodetic reference frame.
In our contribution we will present the first results of parcel-multilooked DS estimation performed within this mixed scatterer framework of the region surrounding Zegveld, The Netherlands based on Sentinel-1 data. Following the parameter estimation in DePSI, we select an IGRS in the region of interest as the InSAR reference point, and transform the relative displacement time series into an absolute frame using the co-located GPS measurements.
References
[1] S. van Asselen, G. Erkens, and F. de Graaf, “Monitoring shallow subsidence in cultivated peatlands,” Proceedings of the International Association of Hydrological Sciences, vol. 382, pp. 189–194, 2020.
[2] G. Erkens, M. J. van der Meulen, and H. Middelkoop, “Double trouble: subsidence and co2 respiration due to 1,000 years of dutch coastal peatlands cultivation,” Hydrogeology Journal, vol. 24, no. 3, pp. 551–568, 2016.
[3] T. Hoogland, J. van den Akker, and D. Brus, “Modeling the subsidence of peat soils in the dutch coastal area,” Geoderma, vol. 171-172, pp. 92 – 97, 2012. Entering the Digital Era: Special Issue of Pedometrics 2009, Beijing.
[4] Y. Morishita and R. F. Hanssen, “Deformation parameter estimation in low coherence areas using a multisatellite insar approach,” IEEE Transactions on Geoscience and Remote Sensing, vol. 53, no. 8, pp. 4275–4283, 2015.
[5] F. van Leijen, Persistent Scatterer Interferometry based on geodetic estimation theory. PhD thesis, TU Delft, 2014.
[6] R. F. Hanssen, “A radar retroreflector device and a method of preparing a radar retroreflector device", U.S. Patent No. 2018236215, 2018.
Estimating the deformation of the Dutch countryside with satellite radar interferometry (InSAR) is notoriously difficult due to rapid soil movement and low coherence [1]. Previous research [2][3][4] has shown that variations in soil moisture can create significant contributions to the observed interferometric phase due to changes in the dielectric properties in the scattering medium. This exacerbates the problem of correctly tracking the deformation of these soft soils, as soil moisture variations can be caused by changes in the ground water level, which is a primary driver of shallow ground deformation in the region [5]. Quantifying and removing the soil moisture phase contribution before phase unwrapping can therefore be a very helpful step to reduce noise in deformation time series.
Employing the phase closure analysis is a very promising approach to estimate the contribution of soil moisture variations without the need for prior phase unwrapping. A set of three SAR images are interfered with each other circularly to form three multilooked interferograms. The summation of the estimated expected values of the three interferometric phases are called the closure phases [6]. Theory states that these closure phases must equal zero for a point scatterer, however, this statement loses its validity when multilooking is employed to form a series of phases over distributed scatterers. These non-zero closure phases have been shown to be caused in part by geophysical processes (i.e. soil moisture variations), and provide us with an opportunity to mitigate geophysical contributions to the wrapped phases [7]. Phase noise as a result of lack of interferometric coherence also contributes to the phase closure, with the added difficulty that large geophysical phase closures go hand-in-hand with low coherences, which makes the phase closure an inherently noisy observable. Therefore, a multilooking strategy which simultaneously suppresses noise but preserves the geophysical closure phases is required. This is accomplished by averaging a large number of pixels over a given region in which ergodicity is assumed.
The vast majority of the Dutch countryside is used for agriculture and is segmented into rectangular parcels surrounded by drainage ditches. With few exceptions, the land cover and groundwater table within a parcel are consistent. This geographical feature provides us with a natural way to segment the scene into multilooked regions which can be described by a few known parameters such as soil type, land cover and groundwater level. The multilooked phases of the distributed scatterers (DS) are condensed to a representative point measurement, which is imported into the Delft Persistent Scatterer Interferometry (DePSI) system [8]. This allows us to accurately estimate and remove atmospheric phase contributions to the measurement and compare the movement of unstable DS points with nearby high-quality point scatterers (PS) in the region.
Closure phases are formed circularly with consecutive 6-day acquisitions from the parcel-based multilooked interferograms. Based on the assumption that noise in closure phases is Gaussian distributed, a numerical model with the inputs of the corresponding coherences of each interferogram and the number of looks is developed in order to perform a significance test. The model output is the standard deviation of the closure phases if solely induced by decorrelation noise. Significance ratios are then computed by dividing closure phases with the numerically simulated standard deviation to estimate the SNR. We find that 1) the closure phases significantly exceed the noise estimate, which suggests a deterministic cause, and 2) there is a promising correlation between the significance ratio and the variation of in-situ groundwater level measurements, which can be used as a proxy for soil moisture variations [9].
This is the first step towards our overall goal of using the observed closure phase to aid in time series analysis. Estimation of soil moisture variation induced interferometric phase differences can subsequently be used to create a phase screen to remove the soil moisture variation component in the observed InSAR phase prior to unwrapping, for instance by using the preliminary interferometric soil moisture model developed by De Zan et al. in [3]. Based on a quantitative correlation analysis between the closure phase observation and theory, we propose recommendations for removing the unwanted soil moisture variation induced phase from an InSAR time series in order to reduce decorrelation and aid in phase unwrapping. This research improves our understanding of the effects of soil moisture variations on the wrapped interferometric phases, and paves the way for deriving a more accurate deformation time series.
References
[1] Y. Morishita and R. F. Hanssen. Temporal decorrelation in l-, c-, and x-band satellite radar interferometry for pasture on drained peat soils. IEEE Transactions on Geoscience and Remote Sensing, 53(2):1096–1104, 2015.
[2] Andrew K Gabriel, Richard M Goldstein, and Howard A Zebker. Mapping small elevation changes over large areas: Differential radar interferometry. Journal of Geophysical Research: Solid Earth, 94(B7):9183–9191, 1989.
[3] Francesco De Zan, Alessandro Parizzi, Pau Prats-Iraola, and Paco López-Dekker. A sar interferometric model for soil moisture. IEEE Transactions on Geoscience and Remote Sensing, 52(1):418–425, 2013.
[4] Simon Zwieback, Scott Hensley, and Irena Hajnsek. Assessment of soil moisture effects on l-band radar interferometry. Remote Sensing of Environment, 164:77–89, 2015.
[5] S. van Asselen, G. Erkens, and F. de Graaf. Monitoring shallow subsidence in cultivated peatlands. Proceedings of the International Association of Hydrological Sciences, 382:189–194, 2020.
[6] Francesco De Zan, Alessandro Parizzi, and Pau Prats. A proposal for a sar interferometric model of soil moisture. In 2012 IEEE International Geoscience and Remote Sensing Symposium, pages 630–633. IEEE, 2012.
[7] Francesco De Zan, Mariantonietta Zonno, and Paco Lopez-Dekker. Phase inconsistencies and multiple scattering in sar interferometry. IEEE Transactions on Geoscience and Remote Sensing, 53(12):6608–6616, 2015.
[8] F. van Leijen. Persistent Scatterer Interferometry Based on Geodetic Estimation Theory. PhD thesis, TU Delft, 2014.
[9] AH de Nijs and RAM de Jeu. Evaluatie van hoge resolutie satelliet bodemvochtproducten met behulp van grondwaterstandmetingen. Stromingen: vakblad voor hydrologen, 28:23–33, 2017.