Identifier
Created
Classification
Origin
08CHENGDU210
2008-09-29 09:49:00
UNCLASSIFIED//FOR OFFICIAL USE ONLY
Consulate Chengdu
Cable title:  

CHENGDU MANIPULATING AIR QUALITY DATA?

Tags:  SENV ENRG ECON PREL CH 
pdf how-to read a cable
VZCZCXRO6721
RR RUEHGH RUEHVC
DE RUEHCN #0210 2730949
ZNR UUUUU ZZH
R 290949Z SEP 08 ZDK
FM AMCONSUL CHENGDU
TO RUEHC/SECSTATE WASHDC 2944
INFO RUEHOO/CHINA POSTS COLLECTIVE
RUCPDOC/DEPT OF COMMERCE WASHINGTON DC
RHMFIUU/DEPT OF ENERGY WASHINGTON DC
RUEAEPA/HQ EPA WASHINGTON DC
RUEHCN/AMCONSUL CHENGDU 3586
UNCLAS CHENGDU 000210 

SENSITIVE
SIPDIS

STATE FOR EAP/CM/SECOR AND HABJAN
STATE FOR OES/WATSON, TALLEY AND THOMPSON STATE PASS TO CEQ CONNAUGHTON DOE FOR INTERNATIONAL/FREDRIKSEN AND BALLOU EPA FOR INTE

E.O. 12958: N/A
TAGS: SENV ENRG ECON PREL CH
SUBJECT: CHENGDU MANIPULATING AIR QUALITY DATA?

REF: A. 2007 BEIJING 07110 B. 2008 BEIJING 00037

UNCLAS CHENGDU 000210

SENSITIVE
SIPDIS

STATE FOR EAP/CM/SECOR AND HABJAN
STATE FOR OES/WATSON, TALLEY AND THOMPSON STATE PASS TO CEQ CONNAUGHTON DOE FOR INTERNATIONAL/FREDRIKSEN AND BALLOU EPA FOR INTE

E.O. 12958: N/A
TAGS: SENV ENRG ECON PREL CH
SUBJECT: CHENGDU MANIPULATING AIR QUALITY DATA?

REF: A. 2007 BEIJING 07110 B. 2008 BEIJING 00037


1. (SBU) This cable contains sensitive but unclassified
information - not for distribution on the Internet.




2. (SBU) Summary: Official data published on the Ministry of
Environmental Protection website strongly suggests that Chengdu
air quality data is being manipulated to appear cleaner than it
is, despite official claims that the data from local air quality
monitors is aggregated and sent to Beijing automatically. Both
the Sichuan provincial-level Environmental Protection Bureau and
the Chengdu Environmental Protection Bureau claim air quality in
Chengdu to be gradually improving, but already in general "very
good." Apparent manipulation going back to at least 2001 makes
it difficult to verify this claim. End Summary.



What Officials Say about the Air

--------------




3. (SBU) The Director of Science and Technology and
International Cooperation at Sichuan's provincial-level
Environmental Protection Bureau, Zhao Lechen, told Congenoff
recently that Chengdu's air quality data is gathered by eight
air quality monitors spread throughout the city. He said that
the monitors are generally placed on top of buildings and
reported data automatically to back to a regional control center
in Chengdu. The data is then passed automatically to Beijing,
where it is posted on the Ministry of Environmental Protection's
website. Zhao asserted that the city's air quality is improving
gradually over time.




4. (SBU) The Deputy Director at Chengdu's Environmental
Protection Bureau, Shen Shaoji, also claimed that data from
local air quality monitoring stations is transmitted
automatically to Beijing. Shen said that moves to reduce
pollution, such as banning coal burning inside the city's third
ring road, restrictions on the use of government vehicles by one
day per week, and encouraging residents who lived less than 1.2
miles from work to avoid driving to work has contributed to the
claimed increase in air quality in the city. Both Shen and Zhao
said that while Chengdu could further improve its air quality,
air quality is in general far better than in cities like Beijing.



Comments

--------------




5. (SBU) Both Shen and Zhao were very optimistic about Sichuan
and Chengdu's environmental conditions and outlook, and neither
told Congenoff that the city faced any significant challenges.
China defines excellent and good air quality as a day having an
air pollution index (API) value between 0 - 100. Analysis of
Chengdu's air quality data between 1 January 2001 and 15
September 2008 shows that during those years there were many
days on which air pollution exceeded an API of 100, but were
recorded as being lower than 100. This year, from 1 January to
15 September, Chengdu recorded 62 days with an API 91-100, but
only 2 days with a value of 101-110. This is consistent with
air quality data manipulation techniques discussed in reftel.




6. (SBU) In a normal distribution of data, one would expect to
see approximately 23 more days with an API 101-110. The same
pattern of data manipulation holds true for previous years. In
2007 Chengdu recorded 77 days between API 91-100, and only 6
between 101-110. For the entire data set from 2001-2008, Chengdu
reported 620 days between API 91-100, about 80 percent more than
would be expected in this random set of pollution data. Days
between 101-110 were less than half of their normal distribution
for a data set.
BOUGHNER