Data Asimilasi Menggunakan Metode Kalman Filter Untuk Mengevaluasi North-American Multi Model Ensemble (NMME) di Indonesia

Adiyatma, Soni (2018) Data Asimilasi Menggunakan Metode Kalman Filter Untuk Mengevaluasi North-American Multi Model Ensemble (NMME) di Indonesia. Undergraduate thesis, InstitutTeknologiSepuluhNopember.

[thumbnail of 1314100050-Undergraduate_Theses.pdf] Text
1314100050-Undergraduate_Theses.pdf - Accepted Version

Download (7MB)

Abstract

Peramalan memiki peran yang sangat penting dalam pengambilan keputusan, terutama dalam upaya penanggulangi bencana alam yang dipengaruhi oleh kondisi geografis dan cuaca. Perlu dilakukan peramalan terhadap cuaca yang akan datang oleh para peneliti dengan menggunakan post-processing dalam upaya memberikan informasi peramalan cuaca yang cepat, akurat, dan reliabel. Peramalan cuaca di Indonesia hanya dapat dilakukan secara terbatas pada lokasi tertentu, yang sangat berbeda dengan data satelit yang tersedia pada posisi grid points. Akan tetapi data tersebut tentunya terdapat bias. Oleh karena itu, post-processing yang melibatkan dua macam data tersebut tidak dapat dilakukan secara langsung namun harus digabungkan dalam upaya representasi data lapangan dan mengurangi bias tersebut. Salah satu metode yang memanfaatkan penggabungan antara model dari suatu keadaan dan data-data pengukuran adalah kalman filter. Dapat disimpulkan bahwa hasil asimilasi dengan menggunakan metode kalman filter memiliki bias yang kecil dibandingkan dengan sebelum dilakukan metode kalman filter. Dapat disimpulkan juga hasil asimilasi setiap musim memiliki bias yang lebih kecil daripada semua musim. Serta dapat disimpulkan bahwa data NMME reliabel di Indonesia dengan model 1, model 2, model 3, dan model 4 menghasilkan MAE antara hasil asimilasi dan data NMME cenderung kecil dan konsisten.
=============== Forecasting plays a very important role in decision making, especially to overcome natural disasters that are affected by geographical and weather conditions. Future weather forecast prediction by researchers used post-processing in order to provide accurate and reliable weather forecasting information. Weather forecasting in Indonesia are limited to a particular location, which is very different from the satellite data that available in the grid points position. However, the data is certainly biased. Therefore, post-processing involved two kinds of data, that can not be done directly but must be combined in to represent field data and reduce the bias. One method that takes advantage of the merging between the model of a state and the measurement data is the kalman filter. It can be concluded that the assimilation result using the Kalman filter method has a small bias compared to the method beside the Kalman filter. Also it can be concluded that assimilation results for each season has a smaller bias than all seasons. It can be concluded that reliable NMME data in Indonesia with model 1, model 2, model 3, and model 4 produce MAE between assimilation and NMME data tends to be small and consistent.

Item Type: Thesis (Undergraduate)
Additional Information: RSSt 519.55 Adi d
Uncontrolled Keywords: Forecast; Indonesia; Grid points; Kalman Filter; PostProcessing.
Subjects: G Geography. Anthropology. Recreation > G Geography (General) > G70.217 Geospatial data
G Geography. Anthropology. Recreation > GA Mathematical geography. Cartography
G Geography. Anthropology. Recreation > GE Environmental Sciences
H Social Sciences > HA Statistics
Q Science > QA Mathematics > QA402.3 Kalman filtering.
Divisions: Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Soni Adiyatma
Date Deposited: 10 Apr 2018 02:33
Last Modified: 25 Sep 2020 02:35
URI: http://repository.its.ac.id/id/eprint/50747

Actions (login required)

View Item View Item