p-ISSN 1693-1246 e-ISSN 2355-3812 July 2019 Jurnal Pendidikan Fisika Indonesia 15 (2) (2019) 107-113 DOI: 10.15294/jpf.v15i2.21937 *Correspondence Address: Kampus Sekaran Gunungpati, Semarang 50229, Indonesia E-mail : aryono_adhi@yahoo.com http://journal.unnes.ac.id/nju/index.php/jpf Invers Modeling Gravity Data for Semi-Infnite Slab Using Matlab A. Adhi 1,2 *, Sismanto 2 , A. Setiawan 2 1 Program Studi Fisika, Universitas Negeri Semarang, Indonesia 2 Program Studi Fisika, Universitas Gadjah Mada, Indonesia Received: 2 March 2019. Accepted: 29 May 2019. Published: 1 July 2019 ABSTRACT Semi-infnite slab modeling has been made through inverse data gravity using Matlab. Inversion of gravity data is done by frst determining the simulation data. Forward modeling uses simulation data to produce an initial guess inversion model expressed with parameters n0 (1), n0 (2), n0 (3) and n0 (4). The forward modeling is performed on the next initial guess that the value of the misft is as small as possible through an iteration using the Jacobian matrix. Accuracy of inversion results is determined by the initial guess and the number of iterations. The results obtained show that inversion modeling is more valid in the inversion modeling process compared to advanced modeling, because the value of the parameters sought is generated from mathematical observations of the observation data. Guesses greatly afect the results of inversions obtained. Initial guesses are given in the form of parameters n0 (1), n0 (2), n0 (3) and n0 (4). The initial guess for the parameters n0 (1), and n0 (2) that are made far deviant does not afect inversion. The initial guess for the parameters n0 (3), and n0 (4) that are made deviating far infuences the inversion caused by a very small RCON value so that the result is NAN. ABSTRAK Telah dibuat pemodelan semiinfnite slab melalui inversi data gravitasi dengan menggunakan Matlab. Inversi data gravitasi dilakukan dengan membandingkan akuisisi data dan perhitungan teoritis. Akuisisi data disusun melalui tahapan forward modeling data sintetik dengan menambahkan error. Model tebakan awal inversi dinyatakan dengan parameter n0(1), n0(2), n0(3) dan n0(4). Nilai misft sekecil mungkin melalui iterasi dengan menggunaan matriks Jacobian antara model tebakan awal dan model recovered. Ketelitian hasil inversi ditentukan oleh tebakan awal dan banyaknya iterasi. Hasil yang didapatkan menunjukkan pemodelan inversi lebih valid dalam proses pemodelan inversi dibandingkan dengan pemodelan maju, karena nilai parameter yang dicari dihasilkan dari olah data observasi secara matematis. Tebakan sangat mempengaruhi hasil inversi yang didapatkan. Tebakan awal diberikan dalam bentuk parameter n0(1), n0(2), n0(3) dan n0(4). Tebakan awal untuk parameter n0(1), dan n0(2) yang dibuat menyimpang jauh tidak mempengaruhi inversi. Tebakan awal untuk parameter n0(3), dan n0(4) yang dibuat menyimpang jauh mempengaruhi inversi yang diakibatkan nilai RCON yang sangat kecil sehingga hasilnya berupa NAN. Keywords: Gravity data; Inversion; Semi-infnite slab. (also called method) (method correction) (Mar- tyshko, Ladovskii, Byzov, & Tsidaev, 2018). The slope flter is an interpretation method used to determine the source near the location based on the potential feld data. (Eshaghza- deh, 2017) The inversion algorithm was deve- loped to simultaneously evaluate the geometry of the error feld and density or depth related parameters of various geological structures in a system of suspended walls, based on ob- stacles. limited electric. because of the obser- INTRODUCTION There several way to solve invers mode- ling data gravity (Supriyadi, 2009). A new gravi- metric data inversion method for a linear prob- lem (reconstruction of the density distribution by a gravitational feld). It is an iterative algo- rithm based on the concept of local correction