Trend analysis of PM2.5 over China

Time series of satellite-based fine
particulate matter (PM2.5) concentrations
over China: 2004-2012
Guannan Geng, Qiang Zhang, Aaron van Donkelaar,
Randall V. Martin, and Kebin He
2013-12-11, San Francisco
Sources and negative effects of fine particles (PM2.5)
Acid rain
Brown Cloud
fine particulate matters in the (ambient)
air with 2.5 micrometers or less in size
Increased morbidity and mortality from both
acute and chronic exposure to ambient fine
particulate matter (PM2.5)
Regional haze is a major concern in China
Primary pollutants in 74 cities
PM2.5 concentration
Heavy PM2.5 pollution occurs in China,
especially in the North China Plain
Beijing haze
How has PM2.5 evolved over the past decade over
• Long-term variation of PM2.5 concentrations is important
for epidemiological studies.
• China has developed control strategies for the
precursors of PM2.5, e.g. SO2 and NOx.
• But ground measurements are limited in China
– Only several research groups have monitored the ground level
PM2.5 concentration before the year 2013.
– National network of PM2.5 monitoring is built up since 2013.
Satellite Remote sensing approach can be used to
enhance our ability to estimate PM2.5 concentration
• Many works have used aerosol optical depth (AOD) to
retrieve PM2.5 concentration
Multiple linear regression
. = ()
Chemical transport model
• In this study, ground-level PM2.5 concentrations for the
period 2004-2012 over China are estimated using CTM
to provide the conversion factor between PM2.5 and AOD.
Satellite-derived surface PM2.5
GEOS-Chem (GC)

van Donkelaar et al, EHP, 2010
van Donkelaar et al, JGR, 2013
Emission trends of the precursors in China
• The MEIC inventory, 2004-2012
Evaluation of model simulations
SNA simulation in the year of 2006
Slope = 0.28, R = 0.32
Slope = 0.26, R = 0.33
Slope = 0.19, R = 0.17
Slope = 0.85, R = 0.39
Slope = 0.44, R = 0.81
Slope = 0.52, R = 0.78
Slope = 0.68, R = 0.71
Slope = 0.86, R = 0.88
Modeled AOD trend compared with MISR
Slope = 0.58, R = 0.66
Slope = 0.79, R = 0.82
Slope = 0.64, R = 0.49
Slope = 0.93, R = 0.74
Satellite-based PM2.5 retrieved from MISR AOD
Trends of the conversion factor and MISR AOD
Modeled PM2.5
Modeled AOD
Modeled PM2.5/AOD
The population weighted national annual mean
PM2.5 trends over China during 2004-2012
PM2.5 trends of some cities in China
The effect of the correlation slope between modeled
and satellite AOD on PM2.5 trend
._ =

• National scale surface PM2.5 concentration derived from MISR
increased from 2004 to 2006 and started to decrease around 20062007, and had a second small peak in 2011. The trend between
2004 and 2012 was +0.5 ug·m-3·yr-1.
• In the Northern China, there are mostly increasing trends, even
above 4 ug·m-3·yr-1 in some regions. While in the Southern China,
PM2.5 decreased in 2012 compared to 2004.
• PM2.5 concentrations in some megacities have started to decrease

similar documents