Bachtiar, Yusra Syarifah and Dr. Eng. Donny Harisuseno, ST., MT. and Jadfan Sidqi Fidari, ST., MT. (2021) Prediksi Laju Infiltrasi Berdasarkan Sifat Porositas Tanah, Distribusi Butiran Pasir, dan Lanau di Sub-DAS Lesti. Sarjana thesis, Universitas Brawijaya.
Abstract
Laju infiltrasi merupakan salah satu informasi penting dalam perencanaan bangunan air, pemetaan, hingga mitigasi bencana. Karakteristik infiltrasi suatu wilayah bersifat dinamis menyesuaikan kondisi wilayah terbaru. Sehingga pembaruan akan informasi terkait harus disesuaikan dan terus dilakukan. Laju infiltrasi dipengaruhi oleh berbagai faktor salah satunya sifat fisik tanah yang meliputi jenis tanah, struktur tanah, sifat porositas tanah, kadar air, konduktivitas hidrolik, kondisi permukaan tanah hingga jenis tutupan lahan serta vegetasi yang dimiliki. Distribusi butiran dan sifat porositas tanah merupakan salah satu faktor yang sangat mempengaruhi kemampuan infiltrasi suatu wilayah, demikian dibuktikan dengan tanah porous yang umumnya memiliki kemampuan infiltrasi yang lebih baik dibandingkan tanah tidak porous. Persamaan empiris akan disusun berdasarkan hubungan laju infiltrasi dengan sifat porositas tanah, distribusi butiran pasir, dan lanau yang kemudian akan disebut sebagai model prediksi. Kedepannya diharapkan persamaan empiris tersebut dapat memudahkan perhitungan laju infiltrasi tanpa membutuhkan pengukuran lapangan agar dapat tercapai efisiensi waktu, tenaga, hingga dana. Persamaan empiris akan didapatkan dari analis regresi berganda dengan tiga variabel bebas yaitu distribusi butiran pasir, distribusi butiran lanau, dan sifat porositas tanah. Data – data variabel bebas tersebut didapatkan dari uji - uji karakteristik tanah yang dilakukan di Laboratorium Air Tanah Fakultas Teknik Universitas Brawijaya. Sedangkan untuk variabel terikat berupa laju infiltrasi konstan didapatkan dari hasil pengkuran lapangan dengan memanfaatkan double ring infiltrometer sebagai alat pengukuran. Data sampel penyusun regresi sebelumnya akan dianalisis dengan analisis korelasi untuk mengetahui kekuatan hubungan masing – masing variabel. Kemudian, beberapa uji akan dilaksanakan demi membuktikan bahwa model prediksi memenuhi BLUE criteria (Best Linier Unbiased Estimator). Rangkaian uji tersebut ialah uji multikolinearitas, normalitas, dan heterokedastisitas sebagai rangkaian analisis asumsi klasik. Sebagai langkah terakhir penggambaran ketepatan model prediksi, dilakukan perhitungan nilai koefisien Nash- Sutcliffe Efficiency (NSE), Root Mean Squared Error (RMSE), Koefisien Korelasi (r) dan Koefisien Determinasi (R2). Pelaksanaan analisis validasi akan memanfaatkan data sampel yang berbeda dari sampel penyusun persamaan regresi. Berdasarkan hasil analisis korelasi yang telah dilakukan, diketahui bahwa hubungan dari laju infiltrasi lapangan dan ketiga variabel bebas memiliki korelasi yang kuat dengan nilai koefisien korelasi >60%. Kemudian terbentuk model prediksi laju infiltrasi seperti berikut: Y = -13,734+1,884x1+0,586x2+1,776x3 (Y = Laju infiltrasi konstan (mm/menit), X1 = Ln distribusi butiran pasir (%), X2 = Ln distribusi butiran lanau (%), dan X3 = Ln porositas tanah (%)). Berdasarkan hasil analisis asumsi klasik dapat disimpulkan bahwa model prediksi memenuhi BLUE criteria dan dapat dikatakan layak untuk digunakan. Hipotesa tersebut dikuatkan oleh hasil analisis validasi yang selaras menyimpulkan bahwa model prediksi memenuhi kriteria – kriteria yang dibutuhkan, dengan nilai NSE=0,84, RMSE=1,13, r=0,79, dan R2 distribusi pasir, distribusi lanau, serta porositas masing – masing 76%, 63%, 63%.
English Abstract
Soil infiltration rate is one of the essential data for waterworks planning, cartography, and mitigation. Infiltration rate's characteristic dynamically changes following the novel condition in the study area. To overcome the mentioned issue, the updated soil infiltration rate information must be constantly made. Soil infiltration rate has various factors that formed their characteristic. To be mentioned, soil physical properties has been cited as the strongest factors including the type of soil, soil structure, soil porosity, water content, hydraulic conductivity, current ground surface condition, and type of land cover. The distribution of granules and the soil porosity is one of the factors that greatly affect the infiltration rate of an area, as proved by porous soils that commonly have a higher infiltration rate than non-porous soils. Empirical equations will be formed base on the relationship of infiltration rate and soil porosity, along with sand and silt content of the soil. This equation will then be referred to as prediction models, and it is expected to be an effective alternative that eases the calculation of infiltration rates without the need for field measurements to achieve efficiency of time, energy, and resources. Empirical equations will be formed with the help of multiple regression analysts. The independent variables consist three variables, there are soil porosity, sand, and silt content of the soil. The three independent variables are obtained from a laboratory test conducted at The Groundwater Laboratory of the Faculty of Engineering, Brawijaya University. The dependent variable of constant infiltration rate is estimated by using the double-ring infiltrometer. All regression’s constituent variables will be analyzed with correlation analysis to determine the strength of each variable's correlation. Then, several tests will be conducted to prove that the prediction model meets the BLUE criteria (Best Linear Unbiased Estimator). The test series will consist of a multicollinearity test, normality test, and heteroskedasticity test as a series of classical assumption analyses. The model prediction would undergo a series of coefficient calculations as a final step in describing the accuracy of the predictive model, the process will be named validation analysis. Nash-Sutcliffe Efficiency (NSE), Root Mean Squared Error (RMSE), Coefficient of Correlation, and Coefficient of Determination (R2) will be used as the validation analysis. The implementation of validation analysis will utilize different sample from sample that used to determine the prediction model. Based on the results of the correlation analysis that has been done, it is known that the relationship of field infiltration rate and the three independent variables have a strong correlation by r>60%. Results of regression analysis formed new model prediction, namely: Y = -13,734+1,884x1+0,586x2+1,776x3 (Y = Constant infiltration rate (mm/min), X1 = Ln form of sand content of soil (%), X2 = Ln form of silt content of soil (%), and X3 = Soil porosity (%)). The classical assumption of prediction model came up with conclussion of meets the BLUE criteria and can be said to be worth using. The hypothesis is validated by the results of a validation analysis concluding that the prediction model meets the required criteria by NSE=0,84, RMSE=1,13, r=0,79. R2 of sand content of soil = 76% and silt content of soil and soil porosity each was 63%.
Other obstract
-
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | 0521070113 |
| Uncontrolled Keywords: | Kata kunci: Laju infiltrasi, distribusi butiran, sifat porositas tanah, analisis regresi, analisis asumsi klasik |
| Subjects: | 600 Technology (Applied sciences) > 627 Hydraulic engineering > 627.5 Reclamations, Irrigation, related topics > 627.52 Irrigation |
| Divisions: | Fakultas Teknik > Teknik Pengairan |
| Depositing User: | yulia Chasanah |
| Date Deposited: | 23 Dec 2021 02:49 |
| Last Modified: | 07 Oct 2024 07:09 |
| URI: | http://repository.ub.ac.id/id/eprint/187593 |
![[thumbnail of YUSRA SYARIFAH BACHTIAR.pdf]](http://repository.ub.ac.id/style/images/fileicons/text.png) |
Text
YUSRA SYARIFAH BACHTIAR.pdf Download (12MB) |
Actions (login required)
 |
View Item |