Fauzi, Fatkhurokhman (2020) Perbandingan Metode Koreksi Bias Dan Statistical Downscaling Pada Data Luaran Earth System Models Untuk Proyeksi Iklim (Studi kasus: Curah Hujan dan Temperatur Maksimum di Indonesia). Masters thesis, Institute Teknologi Sepuluh Nopember.
Preview |
Text
06211750012006-Master_Thesis.pdf Download (5MB) | Preview |
Abstract
Earth System Models (ESM) adalah model yang dapat mensimulasikan, memprediksi perubahan iklim yang terjadi di masa lalu, sekarang, dan membuat skenario perubahan iklim di masa depan. Luaran ESM belum mampu mewakili iklim skala lokal. Statistical Downscaling (SD) adalah proses penurunan skala statis data pada grid skala besar dengan acuan grid skala kecil (skala lokal). Hasil SD masih memiliki bias yang cukup besar, dibutuhkan suatu metode yang berfungsi untuk mengurangi bias. Metode koreksi bias yang digunakan dalam penelitian ini adalah Quantile Delta Mapping (QDM), Analogues with Quantile Mapping Reordering (BCCAQ), dan Inter-Sectoral Impact Model Intercomparison Project (ISIMIP). Penelitian ini menurunkan skala (downscale) dan koreksi bias pada data curah hujan dan temperatur maksimum luaran ESM skenario historical, G4, dan RCP4.5 terhadap data ERA-Interim sebagai proksi pengamatan bersekala lokal. Skill dari metode koreksi bias QDM, BCCAQ, dan ISIMIP pada data skenario historical akan dievaluasi menggunakan diagram Taylor. Skill metode koreksi bias terbaik pada data historical akan diterapkan untuk proyeksi iklim pada data skenario G4 dan RCP4.5. Berdasarkan diagram Taylor ISIMIP memiliki skill yang lebih baik dibandingkan dengan metode koreksi bias QDM dan BCCAQ pada region 1, region 2, dan region 3 serta musiman dan non-musiman. Periode musiman yang memiliki skill terbaik adalah periode musiman September-Oktober-November (SON). Region yang memiliki skill terbaik adalah region 1. Hasil proyeksi iklim pada curah hujan skenario G4 dan RCP4.5 sangat fluktuatif, sedangkan pada temperatur maksimum skenario G4 memiliki temperatur maksimum lebih rendah dibandingkan dengan skenario RCP4.5 pada tahun 2020-2070. Skenario G4 berhasil menurunkan temperatur bumi khususnya di Indonesia pada tahun 2020-2070 sebesar 0-2˚C.
=================================================================================================================================
Earth System Models (ESM) are models that can simulate, predict climate change that occurred in the past, present, and create climate change scenarios in the future. ESM outputs have not been able to represent local-scale climate. Statistical Downscaling (SD) is the process of decreasing the static scale of data on a large scale grid by referring to a small scale grid (local scale). SD results still have a large enough bias, we need a method that works to reduce bias. The bias correction methods used in this study are Quantile Delta Mapping (QDM), Analogues with Quantile Mapping Reordering (BCCAQ), and Inter-Sectoral Impact Model Intercomparison Project (ISIMIP). This study downscale and corrected bias in the rainfall data and the maximum temperature of the ESM output historical, G4, and RCP4.5 scenarios to the ERA-Interim data as a proxy for local-scale observations. The skills of the QDM, BCCAQ, and ISIMIP bias correction methods on historical scenario data will be evaluated using the Taylor diagram. The best bias correction method skills on historical data will be applied to climate projections on G4 and RCP4.5 scenario data. Based on Taylor's diagram, ISIMIP has better skills compared to the QDM and BCCAQ bias correction methods in region 1, region 2, and region 3 as well as seasonal and non-seasonal. The seasonal period that has the best skills is the September-October-November (SON) seasonal period. The region with the best skills is region 1. The results of the climate projection in the G4 and RCP4.5 rainfall scenarios are very volatile, whereas in the maximum temperature the G4 scenario has a lower maximum temperature compared to the RCP4.5 scenario in 2020-2070. The G4 scenario succeeded in reducing the earth's temperature, especially in Indonesia in 2020-2070 by 0-2˚C.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTSt 519.535 Fau p-1 2020 |
| Uncontrolled Keywords: | Koreksi Bias, Downscaling, Diagram Taylor, Proyeksi Iklim, Bias Correction, ESM, Taylor Diagram, Climate Projection. |
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | Fatkhurokhman Fauzi |
| Date Deposited: | 24 Dec 2024 07:44 |
| Last Modified: | 24 Dec 2024 07:44 |
| URI: | http://repository.its.ac.id/id/eprint/74360 |
Actions (login required)
 |
View Item |