Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/164564
Title: Interpolasi Spasial Data Curah Hujan Stasiun Wilayah Danau Toba dengan Ordinary dan External Drift Kriging
Other Titles: Spatial Interpolation of Rainfall Data at Lake Toba Region Stations with Ordinary and External Drift Kriging
Authors: Nurdiati, Sri
Nugrahani, Endar Hasafah
Abisha, Nicholas
Issue Date: 2025
Publisher: IPB University
Abstract: Penelitian ini bertujuan melakukan interpolasi curah hujan di wilayah Danau Toba dengan ordinary kriging (OK) dan external drift kriging (EDK) dengan elevasi (Elev) dan jarak dari danau (JDD) sebagai peubah drift. Data curah hujan dari 34 stasiun (2007–2017) digunakan dengan tiga perlakuan: (1) tanpa imputasi, (2) hanya stasiun dengan data lengkap, (3) semua stasiun dengan imputasi Knearest-neighbors (KNN). Evaluasi dilakukan dengan leave-one-out crossvalidation dengan metrik RMSE, MAE, dan MAPE. Hasil menunjukkan bahwa ketiga metode menghasilkan nilai akurasi yang relatif mirip pada perlakuan 1 dan 3. Namun, EDK_JDD lebih akurat dibandingkan OK dan EDK_Elev pada perlakuan 2 yang menggunakan lebih sedikit stasiun. Dapat disimpulkan bahwa JDD lebih relevan daripada Elev sebagai drift. Perlakuan imputasi KNN tidak meningkatkan akurasi, akan tetapi tetap berguna untuk menjaga kelengkapan data sehingga memungkinkan penggunaan seluruh stasiun dalam proses interpolasi.
This study aims to interpolate rainfall in the Lake Toba region using ordinary kriging (OK) and external drift kriging (EDK) with elevation (Elev) and distance from the lake (JDD) as drift variables. Rainfall data from 34 stations (2007–2017) were used with three treatments: (1) without imputation, (2) only stations with complete data, (3) all stations with K-nearest-neighbors (KNN) imputation. The evaluation was carried out using leave-one-out cross-validation with RMSE, MAE, and MAPE metrics. The results showed that the three methods produced relatively similar accuracy values in treatments 1 and 3. However, EDK_JDD was more accurate than OK and EDK_Elev in treatment 2 which used fewer stations. It can be concluded that JDD is more relevant than Elev as a drift. The KNN imputation treatment did not improve accuracy, but was still useful for maintaining data completeness so that all stations could be used in the interpolation process.
URI: http://repository.ipb.ac.id/handle/123456789/164564
Appears in Collections:UT - Mathematics

Files in This Item:
File Description SizeFormat 
cover_G5401211061_b18559fc702c4172bfc9a76084c4334c.pdfCover1.98 MBAdobe PDFView/Open
fulltext_G5401211061_fea0cc610b4449d386cb9cb32dc3cceb.pdf
  Restricted Access
Fulltext5.96 MBAdobe PDFView/Open
lampiran_G5401211061_b2de22fe1df54e52b31371855912f4fd.pdf
  Restricted Access
Lampiran4.04 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.