Time Series of Land subsidence rate on Coastal Demak Using GNSS CORS UDIP and DINSAR

Most of the studies land subsidence and impact have been done on the north coastal area of Java, especially Semarang and Demak. This landsubsidence has a very serious impact both in the infra structure, the economy and the environment. Techniques of observing landsubsidence using geodetic methods have been widely practiced. Geodetic technology is undergoing rapid development, especially in GNSS navigation satellite technology as well as satellite radar interferometry. Both have advantages and disadvantages of each. In this study will examine the application of DinSAR interferometric techniques and GNSS technology to predict the rate of land subsidence coastal of Demak regency. This paper also emphasize the role of GNSS and DinSAR application in deformation especially for landsubsidence monitoring.


Introduction
Coastal area in Sayung Demak is a very dynamic and have sufficient resource potential big, however, coastal areas are also areas that are very vulnerable to hazards such as flood and inundation land subsidence. The impact of disaster it will be more widespread if not handled properly. Rob floods have caused large loss due to inhibiting economic activity in the Regency Demak. More than 1200 houses flooded and hundreds of residents were forced relocated due to loss of settlement. [1] In this research DinSAR and GNSS measurement were conducted to indentify and analyze the trend of land subsidence. SAR refers to the concept of RADAR using an active sensor that sends radio signals with pulses that bounce off the ground and move back to the sensor. Some of the advantages of using RADAR is that it is not limited by clouds or lack of sunlight. SAR as mentioned, a type of RADAR sensor called Synthetic Aperture Radar, which uses a synthetic antenna, made by a combination of stable data signals (Hanssen, 2001, pp. [10][11]. The most effective way of monitoring ground subsidence can be conducted using dInSAR, GNSS and levelling is to first screen large areas for any movement using dInSAR. [2] The principle of interferometry is to calculate phase differences between SAR images obtained at different times. Since this phase is a measure of distance, this, in theory, represents a change in the height of a point reflected by a signal [3].The problem of SAR image processing became more complex. The complexity is caused by different sources of error Phase comparison without any error correction will produce an inaccurate result, so the result is not realiable and unusable. Recently, different methods have been developed to exclude the error source from phase measurement and hence get much more accurate results. The application of SAR interferometry technology in this research is to monitor landsubsidence deformation. The source of phase error in SAR interferomteri is as follows, Impact of atmospheric change,The impact of orbital inconsistencies, multipath and Topography

SAR Interferometry
the InSAR SAR interferometry, for InSAR stands technique uses the measurement of phase change of signals (interference) between radar images. If there is a point on the surface of the earth moving there is a change in distance between the sensor and the point, it will produce the appropriate phase shift signal. The displacement in phase in the two SAR images measured at each point (pixels) is a phase difference and is known as an interferogram. The phase difference is calculated for each pixel. When InSAR is used to recognize and measure ground deformation, this process is referred to as the InSAR differential [4].
The general formation of differential Interferrograms shown in figure 1.

GNSS Cors
Use of Global Navigation Satellite The GNSS system (GNSS) is increasing with accuracy reaches centimeters for relative position using the CORS station [5]. CORS applications include can be used for geoid modeling purposes (Wu et al., 2015) and for deformation purposes Dam [6].GNSS, like dInSAR, uses phase measurements from satellites. There are several types of GNSS measurements and many different systems included in GNSS terms, such as GPS, GLONASS, Galileo among others [7] The most accurate type of measurement is the relative static position. Here there are several recipients used by being placed above the point where they are static during the session. To get a GNSS position, the receiver must get a selection to at least four different satellites, this is to get the four pseudorange data together so that it gets a 3D position and a clock error [7]. [8] describe in more detail about how GNSS measurements are performed in "GPS Satellite Surveys". The GNSS CORS UDIP station was developed at the Diponegoro University Department of Geodesy in December 2012. The GNSS CORS UDIP installation is on the 3rd floor, while the server is in the laboratory Geodetic Department Measurement and Mapping,Faculty of Engineering, Diponegoro University. Installation / installation of the GNSS system [9] This CORS UDIP is based on needs survey needs, geodetic mathematical models, and digital signal processing. Description from the CORS UDIP GNSS station can be seen Figure 2.

Data and Method
We have been conducted several research investigations to this area study by field observations by (mapping the landsubsidence area, pictures, conduct GPS measurement and InSAR analysis). In this research there are two pairs of images were processed in two year 2016 and 2017. In this case, the subsidence is derived using differential method, and data SRTM was used for generating the DEM for the area. In the interferogram, the phase difference is described as the distance difference measured by the line of sight (LOS) radar and includes topography, orbit shifts, surface deformation and atmospheric effects [10]. The Sentinel Data were used shown in table 1, including data acquisition, polarisation and direction scanning.   The further shown that land subsidence affects north parts of Sayung. This is particularly true for the northwestern districts of the city and situated north of the boundary at alluvial sediments. In the northern portion of the researched area, single points characterized by rapid subsidence are probably connected to tidal flood but have still to be checked in the field. The two interferometric methods determined similar results with small differences. The deformation fringes are to be verified in a future research article by advanced multitemporal DInSAR, and next research must be used by precise levelling measurements. Results from DInSAR processing shown in Figure 4 shown sayung district tend to subsided with a max value of 4.7. cm . The magnitude of the decline varies in every region of In Figure 4 shown the Phase unwrapping data, the next step is to change the phase to shift to determine the Los quantity. The result of the phase change process to the landsubsidece. DInSAR processing shown in Figure 5 show sayung district tend to subsided with a max value of 19.5cm in 2016. Figure 5. The gps data were processed using the GAMIT 10.6 [12]. GLRED / GLOBK processing can be done after the GAMIT process that produces h-files which is the main requirement in processing GLOBK. The results obtained from this processing are 3D cartesian coordinates (X, Y, Z) and topocentric coordinates (N, E, U) with their respective deviation. To obtain the height value obtained from the difference between two high values of different epoch observations. List of topografi coordinat in 2018 can be seen in table 5 The standard deviations of GPS-derived relative ellipsoidal heights in 2018 from all surveys were in general better than 3mm -4cm (see at Table 5). A few points have slightly larger standard deviations, due to the lack of observed data caused by the signal obstruction Analysis of Land Deformation period 2015-20108 period in the coastal area can be seen in Table 6. Results from GPS survye show that vertical displacement in Demak has spatial variations, ranging from 0.8 to 17,91 cm/year (see Table 6

Closing Remark
One of the weakness of DinSAR Processing, is that the coherence is still quite low, this is due to temporal decorrelation, so there is quite a lot of data with adjacent temporal and baseline distances. For temporal distance data acquisition can be done every 6 months with a baseline distance = 150m. From the results shown there are still have low coherence due to temporal, atmospheric and noise decorrelation, it is necessary to increase the accuracy using the PS InSAR technique. In this study we propose , It should be installed a stable BM for a long time and validate the DInSAR and GNSS analysis with leveling measurements, which aim to improve understanding of the characteristics of the geohazard process and enable a good correlation between methods.