Time Series Forecasting of Air Quality Based On Regional Numerical Modeling in Hong Kong

Journal of Geophysical Research D: Atmospheres - Tập 123 Số 8 - Trang 4175-4196 - 2018
Tong Liu1, Alexis K.H. Lau2,1, Kai Sandbrink3, Jimmy Chi Hung Fung4,1
1Division of Environment and Sustainability, Hong Kong University of Science and Technology, Hong Kong
2Civil and Environmental Engineering, Hong Kong University of Science and Technology, Hong Kong
3Institute of Neuroinformatics, ETH Zurich, Zurich, Switzerland
4Department of Mathematics, Hong Kong University of Science and Technology, Hong Kong

Tóm tắt

AbstractBased on prevailing numerical forecasting models (Community Multiscale Air Quality [CMAQ] model , Comprehensive Air Quality Model with Extensions, and Nested Air Quality Prediction Modeling System) and observations from monitoring stations in Hong Kong, we employ a set of autoregressive integrated moving average (ARIMA) models with numerical forecasts (ARIMAX) to improve the forecast of air pollutants including PM2.5, NO2, and O3. The results show significant improvements in multiple evaluation metrics for daily (1–3 days) and hourly (1–72 hr) forecast. Forecasts on daily 1‐hr and 8‐hr maximum O3are also improved. For instance, compared with CMAQ, applying CMAQ‐ARIMA reduces average root‐mean‐square errors (RMSEs) at all stations for daily average PM2.5, NO2, and O3in the next 3 days by 14.3–21.0%, 41.2–46.3%, and 47.8–49.7%, respectively. For hourly forecasts in the next 72 hr, reductions in RMSEs brought by ARIMAX using CMAQ are 18.2% for PM2.5, 32.1% for NO2, and 36.7% for O3. Large improvements in RMSEs are achieved for nonrural PM2.5and rural NO2using ARIMAX with three numerical models. Dynamic hourly forecast shows that ARIMAX can be applied for forecast of 7‐ to 72‐hr PM2.5, 4‐ to 72‐hr NO2, and 4‐ to 6‐hr O3. Besides applying ARIMAX for NO2, we recommend a mixed forecast strategy to ARIMAX for normal values of PM2.5and O3and employ numerical models for outputs above 75th percentile of historical observations. Our hybrid ARIMAX method can combine the advantage of ARIMA and numerical modeling to assist real‐time air quality forecasting efficiently and consistently.

Từ khóa


Tài liệu tham khảo

10.1007/BF02294359

10.1175/JAM2531.1

10.1029/2001JD001409

10.1016/0020-0190(87)90114-1

Box G. E., 1970, Forecasting and control, Time Series Analysis, 3, 75

10.1115/1.2128636

10.1007/s00703-003-0070-7

10.1080/10937404.2014.946164

10.1029/2005JD006311

Diebold F. X., 1995, Comparing predictive accuracy, Journal of Business & Economic Statistics, 253, 10.1080/07350015.1995.10524599

10.1016/j.atmosenv.2009.11.007

ENVIRON(2014).CAMx users' guide version 6.1. Retrieved fromhttp://www.camx.com/files/camxusersguide_v6-10.pdf

10.1111/j.1539-6924.2011.01630.x

10.1016/j.atmosenv.2011.10.020

10.5194/acp-7-4639-2007

10.1016/j.atmosenv.2005.11.041

10.5194/acp-6-3181-2006

10.1029/2007JD008761

Huang S., 2015, Application of ensemble forecast and linear regression method in improving PM10 forecast over Beijing areas, Acta Scientiae Circumstantiae, 35, 56

10.1016/j.atmosenv.2010.03.017

10.1029/2008JD010151

Khandakar Y., 2008, Automatic time series forecasting: the forecast Package for R, Journal of Statistical Software, 27

10.1007/s00477-004-0228-y

10.1016/j.atmosenv.2009.06.039

10.1080/10473289.2004.10470949

10.1007/s00477-009-0361-8

10.1016/0304-4076(92)90104-Y

10.1016/j.rser.2008.02.002

10.5194/acp-17-935-2017

10.1016/j.atmosenv.2009.01.055

10.1029/2005JD006837

Louie P. K., 2005, Seasonal characteristics and regional transport of PM2.5 in Hong Kong, Atmospheric Environment, 39, 1695

10.1016/j.chemosphere.2004.10.032

10.1016/j.envres.2003.11.003

10.1016/j.chemosphere.2007.09.050

Madronich S. Flocke S. Zeng J. Petropavlovskikh I. &Lee‐Taylor J.(2003).Tropospheric Ultraviolet and Visible (TUV) radiation model. Retrieved fromhttps://www2.acom.ucar.edu/modeling/ tropospheric‐ultraviolet‐and‐visible‐tuv‐radiation‐model.

10.1029/2005JD005858

10.1016/1352-2310(94)90083-3

10.1016/1352-2310(94)90084-1

10.1155/2012/894714

10.1016/j.envpol.2004.11.010

10.4236/gep.2015.38002

Pijnappel A., 2011, Statistical post processing of model output from the air quality model LOTOS‐EUROS

10.1016/j.atmosenv.2011.12.052

Pleim J. E.(2006).A new combined local and non‐local PBL model for meteorology and air quality modeling. In5th Annual CMAS Models‐3 User's Conference Chapel Hill NC October(pp. 16–18).

10.1007/0-306-47460-3_30

10.1016/S0377-2217(99)00069-7

10.1016/j.atmosres.2013.05.021

Roselle S. &Binkowski F.(1999).Cloud dynamics and chemistry.Science Algorithms of the EPA Models‐3 Community Multiscale Air Quality (CMAQ) Modeling System. Tech. Rep. EPA/600/R‐99/030 EPA.

Scholz F. W., 1985, Encyclopedia of Statistical Sciences, 1

10.1016/1352-2310(95)00321-5

10.1029/1999JD900423

Sparse Matrix Operator Kernel Emissions v3.5 User's Manual(2013). Retrieved fromhttps://www.cmascenter.org/smoke/documentation/3.5/html/

10.1016/j.atmosenv.2007.10.073

10.1029/JD095iD10p16343

10.1029/2010JD014040

10.1023/A:1013833217916

Weather Research Forecast Users' Guide Version 3.6(2014). Retrieved fromhttp://www2.mmm.ucar.edu/wrf/users/docs/ user_guide_V3/

10.1029/2006JD007598

10.1002/joc.2419

10.1016/j.atmosenv.2012.06.071

Wooldridge J. M., 2015, Introductory econometrics: A modern approach

10.1029/2008JD009991

10.1007/s00703-011-0179-z

Yarwood G. Rao S. Yocke M. A. &Whitten G. Z.(2005).Updates to the carbon bond chemical mechanism: CB05. Final Report to the US EPA RT‐0400675 8.

10.1002/asl.125

10.1016/j.atmosenv.2005.11.030

10.1029/1999JD900876

10.5194/acp-3-2067-2003

10.1016/j.atmosenv.2012.06.031

10.1016/j.atmosenv.2012.02.041

10.1016/j.atmosenv.2015.07.022

Zhao H. Wu Z. L. Xie Y. Y. Bai Z. P. &Zhang C. (2007).The experiments of modifying the CMAQ products by using the CMAQ‐MOS and the CMAQ‐ANNs. The 6th Annual CMAS Conference Chapel Hill NC.

10.1016/j.atmosenv.2009.04.060

Thông tin
Thông tin xuất bản

Journal of Geophysical Research D: Atmospheres

Tập 123 Số 8

4175-4196

Thông tin tác giả