Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/96612
標題: 利用函數型資料分析校準氣象預測
Calibrating Weather Forecasting by Functional Data Analysis
作者: 汪欣樺
Xin-Hua Wang
關鍵字: 空間資料;長期資料;函數型資料分析;氣象預測;Spatial data;Longitudinal data;Functional data analysis;Weather forecasting
引用: [1] BBC. India heatwave toll passes 1,000, 2015. URL http://www.bbc.com/news. [2] Fang Yao, Hans-Georg Müller, and Jane-Ling Wang. Functional data analysis for sparse longitudinal data. Journal of the American Statistical Association, 100(470):577–590, 2005. [3] Lu-Hung Chen and Ci-Ren Jiang. Multi-dimensional functional principal component analysis. Statistics and Computing, 27(5):1181–1192, 2017. [4] Matthew Avery. Literature review for local polynomial regression, 2013. URL http://www4.ncsu.edu/ mravery/AveryReview2.pdf. [5] Vadim Zipunnikov, Brian Caffo, David M Yousem, Christos Davatzikos, Brian S Schwartz, and Ciprian Crainiceanu. Functional principal component model for high-dimensional brain imaging. NeuroImage, 58(3):772–784, 2011. [6] Yulia Gel, Adrian E Raftery, and Tilmann Gneiting. Calibrated probabilistic mesoscale weather field forecasting: The geostatistical output perturbation method. Journal of the American Statistical Association, 99(467):575–583, 2004. [7] 黃敏嘉. 多變量函數-函數型線性迴歸. 中興大學統計學研究所學位論文, 2017. [8] 自由時報電子報. 寒流來襲4 天全台154 人猝死, 2017. URL http:// www.ltn.com.tw/. [9] 中央氣象局全球資訊網. URL http://www.cwb.gov.tw/V7/. [10] 行政院環境保護署環境資源資料開放平台. URL https:// opendata.epa.gov.tw/.
摘要: 
近年來隨著科技的進步,氣象預報的準確度已逐年提升。然而受到計算資源
與觀測技術的限制,氣象科學界仍難以提供高解析度且長時間的精準氣象預報。
本篇論文主要目的是要利用氣象觀測資料來校準長時間的氣象預報,並將其解析
度提升至鄉鎮解析度。氣象是隨著時間及空間的變化而改變,在不同時間和空間
觀測到的資料,可將其視為函數型資料。本篇論文假設氣象觀測資料及預測資料
來自兩個有相關性的隨機函數(random function),並考慮每個隨機函數中空間及
時間的相關性,再利用函數型資料分析(Functional Data Analysis) 的技巧,建立
出校準氣象預測的統計模型。

Recently, weather forecasts become accurate as technology advances. However,
high-resolution long-term forecasts are still challenging due to the difficulty
of measurement certain atmospheric parameters (e.g. soil moisture) and the limitation
of computation resource. In this article we focus on the calibration of longterm
weather forecasts by historical weather observations and predictions. The
atmospheric parameters are treated as continuous spatial-temporal functions, and
function-on-function linear regression models are utilized. Our experiment on temperature
data in Taiwan shows that our achieves better calibration results compared
to state-of-the art approach.
URI: http://hdl.handle.net/11455/96612
Rights: 同意授權瀏覽/列印電子全文服務,2017-08-23起公開。
Appears in Collections:統計學研究所

Files in This Item:
File SizeFormat Existing users please Login
nchu-106-7104018016-1.pdf4.21 MBAdobe PDFThis file is only available in the university internal network    Request a copy
Show full item record
 

Google ScholarTM

Check


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