Hakim, Muhammad Lukman (2018) Prediksi Cuaca Harian Dengan Bayesian Model Averaging Untuk Antisipasi Bencana Hidrometeorologi. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211440000038-Undergraduate_Theses.pdf Download (2MB) | Preview |
Abstract
Indonesia dalam beberapa tahun terakhir mengalami fenomena perubahan iklim. Salah satu dampak darinya yaitu terjadinya bencana hidrometeorologi. Badan Meteorologi, Klimatologi, dan Geofisika atau BMKG dibentuk untuk melaksanakan tugas pemerintahan di bidang Meteorologi, Klimatologi, dan Geofisika. Perlu upaya untuk meningkatkan kualitas dari informasi cuaca yang diberikan. Pada penelitian ini dilakukan kalibrasi prediksi ensemble suhu maksimum, suhu minimum, dan kelembaban menggunakan Bayesian Model Averaging dengan memanfaatkan data observatif stasiun pengamatan BMKG dan data luaran model NWP. Prediksi anggota ensemble yang digunakan berasal dari model PLS, PCR, Regresi Ridge, dan Stepwise Regression. Hasil yang didapatkan prediksi anggota ensemble menghasilkan hasil yang cukup akurat dan masuk kategori sedang-baik. Nilai RMSE yang didapat menujukan bahwa prediksi model PLS dan Regresi Ridge lebih baik daripada dua model lainnya. Sementara hasil BMA menunjukkan bahwa BMA mampu mengatasi kasus underdispersive (nilai prakiraan terpusat pada suatu nilai dengan varians yang rendah) dan ketidakpastian pada prediksi raw ensemble, serta BMA menghasilkan prediksi yang lebih akurat. Selain itu disimpulkan pula bahwa perlu adanya pendekatan statistik dalam memanfaatkan luaran NWP untuk prediksi cuaca.
======================================================================================================
Indonesia in the last few years has experienced climate change phenomenon. One of the impacts of this is the occurrence of hydrometeorological disaster. BMKG was formed to carry out government duties in the field of Meteorology, Climatology, and Geophysics in accordance with the provisions of applicable legislation. It is necessary to improve the quality of weather information provided. In this research, calibration of ensemble prediction of maximum temperature, minimum temperature, and humidity using Bayesian Model Averaging by using observational data of BMKG observation station and NWP outcomes. The ensemble member predictions used are derived from PLS, PCR, Ridge, and Stepwise Regression models. The results obtained predicted MOS produce results that are quite accurate and into the category of medium-well. The obtained RMSE values indicate that the prediction of PLS and Ridge Regression models is better than the other two models. While the BMA results show that BMA is able to overcome the underdispersive case and uncertainty in predictions of raw ensemble, and BMA produces a more reliable prediction. It is also concluded that statistical approaches in utilizing NWP outcomes for weather prediction are required
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSt 519.542 Hak p-1 3100018076806 |
| Uncontrolled Keywords: | Bayesian Model Averaging, BMKG, Cuaca Harian, Numerical Weather Prediction, Daily Weather |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression Q Science > QA Mathematics > QA278.5 Principal components analysis. Factor analysis. Correspondence analysis (Statistics) Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory. |
| Divisions: | Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Muhammad Lukman Hakim |
| Date Deposited: | 31 Oct 2020 05:55 |
| Last Modified: | 31 Oct 2020 05:55 |
| URI: | http://repository.its.ac.id/id/eprint/55972 |
Actions (login required)
 |
View Item |