Analisis Perbandingan Pola Strain Rate pada Sesar Lasem, Semarang, Kendeng Menggunakan Metode Interpolasi Radial Basis Function dan Kriging Berbasis Integrasi InSAR-GNSS
Dhia Olivia Farmalinda, Cecep Pratama, S.Si., M.Si., D.Sc.
2026 | Skripsi | TEKNIK GEODESI
Zona pertemuan Sesar Lasem, Semarang dan Kendeng (LSK) merupakan area dengan aktivitas sesar yang berpotensi menimbulkan gempa maupun deformasi. Diperlukan monitoring strain rate berkelanjutan melalui integrasi InSAR-GNSS untuk mengatasi keterbatasan kedua metode. Proses integrasi membutuhkan metode interpolasi, baik dengan fungsi basis spasial atau metode geostatistik berbasis pembobotan tetangga. Radial Basis Function (RBF) unggul dalam menangani permukaan halus dengan implementasi sederhana, namun tidak mempertimbangkan struktur korelasi spasial. Sebaliknya, kriging memanfaatkan struktur spasial melalui variogram dan mampu menghasilkan estimasi ketidakpastian, tetapi memerlukan pemodelan variogram yang lebih kompleks. Perbedaan karakter keduanya berpotensi memengaruhi nilai dan pola strain rate yang dihasilkan, namun perbandingan pengaruhnya terkhusus di zona LSK hingga saat ini belum banyak dikaji.
Data yang digunakan pada periode 2022 s.d. 2024, terdiri dari InSAR yang diolah dengan LiCSBAS dan data GNSS sekunder berupa nilai laju pergeseran dari stasiun CORS dan campaign. Nilai laju pergeseran hasil integrasi kemudian masing-masing digunakan untuk estimasi strain rate menggunakan algoritma Velocity Interpolation for Strain Rate (VISR). Nilai yang diperoleh juga dilakukan uji signifikansi dan visualisasi pola deformasi yang terbentuk.
Nilai laju pergeseran pada titik sampel dengan metode interpolasi RBF sebesar -16,8 s.d. 27,2 mm/tahun (EW) dan -13,3 s.d. 2,2 mm/tahun (NS), sedangkan metode kriging berkisar antara -3,5 s.d. 22,5 mm/tahun (EW) dan -0,3 mm/tahun s.d. 4,1 mm/tahun (NS). Nilai strain rate dari laju pergeseran hasil metode RBF menghasilkan principal strain rate pada komponen ekstensi mencapai 373,9 nstrain/tahun dengan rata-rata simpangan baku 79,9 nstrain/tahun, sedangkan komponen kompresi mencapai -594,5 nstrain/tahun dengan rata-rata simpangan baku 75,2 nstrain/tahun. Persentase titik signifikan kedua komponen sebesar 29,17?n 50%. Nilai dilatation rate berada pada -540,8 s.d 352,6 nstrain/tahun dengan rata-rata simpangan baku 108,2 nstrain/tahun, diperoleh 20,83% titik signifikan. Sementara itu, maximum shear strain rate memiliki nilai berkisar 30,2 s.d. 458,8 nstrain/tahun dengan rata-rata simpangan baku 57,9 nstrain/tahun, diperoleh 54,17% titik signifikan. Hasil berbasis kriging menghasilkan nilai principal strain rate komponen ekstensi mencapai 422,1 nstrain/tahun dengan rata-rata simpangan baku 55,9 nstrain/tahun, sedangkan komponen kompresi mencapai -382 nstrain/tahun dengan rata-rata simpangan baku 49,6 nstrain/tahun. Persentase titik signifikan kedua komponen sebesar 50?n 54,17%. Nilai dilatation rate pada rentang -375,2 s.d 361,3 nstrain/tahun dengan rata-rata simpangan baku 74,7 nstrain/tahun, diperoleh 40,83% titik signifikan. Nilai maximum shear strain rate pada rentang 5,7 s.d 359,7 nstrain/tahun dengan rata-rata simpangan baku 38,4 nstrain/tahun, diperoleh 70,83% titik signifikan. Pola pada metode kriging menghasilkan orientasi lebih teratur dengan kompresi meluas di sekitar Sesar LSK, sedangkan RBF menunjukkan nilai regangan geser lebih tinggi di wilayah yang sama.
The Lasem, Semarang, and Kendeng (LSK) fault zone is an area with fault activity that has the potential to cause earthquakes and deformation. Continuous strain rate monitoring through InSAR-GNSS integration is required to address the limitations of both methods. The integration process requires an interpolation method, either using spatial basis functions or geostatistical methods based on neighbor weighting. The Radial Basis Function (RBF) excels at handling smooth surfaces with simple implementation, but it does not account for spatial correlation structures. In contrast, kriging utilizes spatial structure through a variogram and is capable of producing uncertainty estimates, but requires more complex variogram modeling. The differences in their characteristics have the potential to influence the resulting strain rate values and patterns, however a comparison of their specific effects in the LSK zone has not been extensively studied to date.
The data used for the 2022–2024 period consist of InSAR processed using LiCSBAS and secondary GNSS data in the form of displacement rates from CORS stations and field campaigns. The integrated displacement rates were then used to estimate strain rates using the Velocity Interpolation for Strain Rate (VISR) algorithm. The obtained values were also subjected to significance tests, and the resulting deformation patterns were visualized.
The displacement rates at the sample points, calculated using the RBF interpolation method, ranged from -16.8 to 27.2 mm/year (EW) and from -13.3 to 2.2 mm/year (NS), while the kriging method ranges from -3.5 to 22.5 mm/year (EW) and from -0.3 mm/year to 4.1 mm/year (NS). The strain rate values from the RBF method yielded a principal strain rate in the extension component of 373.9 nstrain/year with a standard error of 79.9 nstrain/year, while the compression component reached -594.5 nstrain/year with a standard error of 75.2 nstrain/year. The percentage of significant points for the two components was 29.17% and 50%, respectively. The dilatation rate ranged from -540.8 to 352.6 nstrain/year with a standard error of 108.2 nstrain/year, yielding 20.83% significant points. Meanwhile, the maximum shear strain rate ranged from 30.2 to 458.8 nstrain/year with a standard error of 57.9 nstrain/year, yielding 54.17% significant points. The kriging-based results showed that the principal strain rate for the extension component reached 422.1 nstrain/year with a standard error of 55.9 nstrain/year, while the compression component reached -382 nstrain/year with a standard error of 49.6 nstrain/year. The percentage of significant points for both components was 50% and 54.17%, respectively. The dilatation rate values ranged from -375.2 to 361.3 nstrain/year with a standard error of 74.7 nstrain/year, yielding 40.83% significant points. The maximum shear strain rate values range from 5.7 to 359.7 nstrain/year with a standard error of 38.4 nstrain/year, yielding 70.83% of significant points. The patterns obtained using the kriging method show a more regular orientation with widespread compression around the LSK Fault, while the RBF method indicates higher shear strain values in the same region.
Kata Kunci : Sesar LSK, InSAR-GNSS, RBF, Kriging, Principal strain rate, Dilatation rate, Maximum shear strain rate
