Kalibrasi Prediksi Suhu Dan Kelembapan Jangka Pendek Menggunakan Ensemble Model Output Statistics

., Fachrunisah (2017) Kalibrasi Prediksi Suhu Dan Kelembapan Jangka Pendek Menggunakan Ensemble Model Output Statistics. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
1313100008-Undergraduate_Theses.pdf - Published Version

Download (3MB) | Preview

Abstract

Badan Meteorologi, Klimatologi dan Geofisika dalam perkembangannya telah melakukan peramalan cuaca dengan menggunakan Numerical Wheather Prediction (NWP). Akan tetapi, NWP masih mempunyai kekurangan, yaitu prediksi yang dihasilkan bias pada wilayah yang sempit. Sehingga dilakukan post-processing dengan menggunakan Model Output Statistics (MOS). MOS yang mampu mengatasi multikolineriatas adalah dengan pendekatan principal component regression, stepwise regression dan SIMPLS. Pada penelitian ini menghasilkan prediksi untuk temperatur maksimum, minimum dan kelembapan udara dengan pendekatan principal component regression, stepwise regression dan SIMPLS. Dari ketiga model tersebut dilakukan kalibrasi ensemble dengan menggunakan ensemble model output statistics (EMOS). Hasil penelitian menunjukkan bahwa model stepwise regression memiliki kontribusi paling besar dalam model kalibrasi EMOS dibandingkan dengan SIMPLS dan principal component regression. Model EMOS juga baik dalam memprediksi temperatur dan kelembapan udara jika dibandingkan dengan prediksi NWP karena hasil prediksi EMOS memiliki nilai CRPS 0,405 dan 0,395 untuk temperatur maksimum dan minimum. ======================================================================================================================== Nowadays, Department of Meteorology, Climatology and Geophysics has conducted weather forecasting by using NWP. But, in case, NWP still has a weakness. The prediction produced are generally biased on the narrow region. So, it's needed to do post-processing by using Model Output Statistics. Model Output Statistics solves multicolyneriaty with the approachment of Principal Component Regression, Stepwise Regression and Statistically Inspired Modification Partial Least Square. This research will produce the prediction of maximum and minimum temperature and relative humidity with Principal Component Regression, Stepwise Regression and Statistically Inspired Modification Partial Least Square. That three models will be calibrated with ensemble calibration by using Ensemble Model Output Statistics (EMOS). The final analysis found that the stepwise regression model has the greatest contribution in the EMOS calibration model compared to SIMPLS and principal component regression. The EMOS model is also good at predicting temperature and relative humidity when compared to NWP predictions because the EMOS prediction results have CRPS values 0,405 and 0,395 for maximum and minimum temperatures.

Item Type: Thesis (Undergraduate)
Additional Information: RSSt 519.542 Fac k
Uncontrolled Keywords: Ensemble Model Output Statistics; Kelembapan udara; Model Output Statistics; NWP; Temperatur; Ensemble Model Output Statistics, Model Output ; Statistics; NWP; Relative Hu madity; Temperature
Subjects: Q Science > QA Mathematics > QA278.2 Regression Analysis
Q Science > QA Mathematics > QA278.5 Principal components analysis. Factor analysis. Correspondence analysis (Statistics)
Divisions: Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Fachrunisah .
Date Deposited: 17 Jan 2018 04:41
Last Modified: 05 Mar 2019 04:02
URI: http://repository.its.ac.id/id/eprint/48531

Actions (login required)

View Item View Item