1	OlympicWatch Following the course of air pollution in Beijing, Summer 2008
2	"Kenneth A." Kenneth A. Rahn Graduate School of Oceanography University of Rhode Island Narragansett, RI 02882-1197 July–October 2008
3	"Introduction" Introduction Data Discussion Comments and Predictions
4	This little PowerPoint presentation is designed to keep readers up to date on the progress of Chinese efforts to reduce air pollution to acceptable levels during the 2008 Summer Olympics in Beijing.
5	I will update it daily (or so) as we approach the Olympics and pass through it.
6	It reports official figures on the daily Air Pollution Index (API) for Beijing and displays them graphically.
7	The API for Beijing is normally set by the levels of PM₁₀ (particulate matter, or particles, smaller than 10 μm in diameter).
8	The daily concentrations of PM₁₀ from 10–20 measuring stations are first averaged over the city, and then transformed to an Air Pollution Index (API).
9	The critical value of the API is 100, which corresponds to the borderline (according to the Chinese) between healthful and unhealthful air.
10
11	APIs of 100 or less are called “blue sky” days. Beijing pledged to keep the API’s in this range as a condition for getting the Olympics.
12	Now the time has come to see whether they can actually do it. So far, two weeks before the Olympics, they have not.
13	The first graph below shows the APIs leading up to the Olympics, from July 1^st on.
14
15	These data come from the Chinese Ministry of Environmental Protection http://www.mep.gov.cn/quality/air.php3 (Archived data can be searched at the bottom of the page.)
16	It shows the healthful and unhealthful ranges, and the beginning of the big ban (cars, construction, spray-painting, industrial emissions, refining, quarrying, etc.) on July 20^th.
17	It is clear that nothing changed on the 20^th. Three days afterward, the pollution levels began to increase sharply, and ultimately doubled.
18	Then on the 28^th and 29^th they fell just into the healthful range (96 and 90), and on the 30^th down to 43. The next day, the situation worsened slightly (69), and then on August 1^st the best air since before July (27) appeared.
19	These strong cycles illustrate the magnitude of the task that the Chinese government faces.
20	Does the longer period of data show a decrease attributable to the emergency measures? To address this question (highly preliminarily), I constructed a three-day running average from July 1^st through August 8^th(now updated to the 10^th). That smoothes out some of the variations and better reveals longer-term trends. The result is given in the next slide. It shows no detectable trend in the peaks, but two lower minima in August.
21	Aug 19^th (Tuesday): We now have enough data to better evaluate the plot of running averages. It seems to be divided into July vs. August, both of which have short-period and long-period cycles. At the beginning of August, the whole pattern dropped down by 20–30 API units. I first thought this was a downward trend during the first half of August, but it now appears to be more of a step function. It is more consistent with meteorological changes than with control efforts, which if they had kicked in rapidly, would have done so on July 20^th rather than two weeks later. Thus I see this pattern as arising more from Mother Nature than from pollution controls.
22
23	The second graph shows the ranges of API, expressed as maximum urban, minimum urban, maximum outside, and minimum outside.
24	“Outside” begins near urban Beijing and extends to 100 km (70 miles) out.
25
26	These data come from the Beijing Environmental Protection Bureau http://www.bjepb.gov.cn/air2008/Air.aspx?time=2008-7-15 (Dates at the end can be changed as needed.)
27	6 October 2008: General observations on the API during summer and fall of 2008 There was no obvious change when the ban went into effect on 20 July. In fact, the API shortly thereafter went into its worst episode of the summer. There was no obvious change when the ban terminated on 20 September. In fact, the API shortly thereafter plummeted to its lowest values of the summer. Observations 1 and 2 are the opposite of what would be expected from banning sources and then releasing them. The lowest daily lows in Beijing did not change perceptibly over the entire period of July into October. The highest highs outside Beijing City were virtually identical to the values inside the city. Observations 1 through 5 are consistent with regional control of Beijing’s air quality, and inconsistent with a detectable effect from the ban. Obviously, the ban had an effect on Beijing’s air quality. But it cannot be seen from the record of APIs.
28	The regionality of the pollution is shown by (1) the persistent similarity of the city and outside maxima, and (2) the similarity of the city and outside minima during the highest concentrations.
29	Note also that at the peak of the long episode in late July, EVERY monitoring station was over 100. That includes sites out to 70 miles (100 km) from Beijing.
30	This shows the regionality of Beijing’s problem very dramatically.
31	In other words, the periods of highest concentrations are regional, and affect everywhere around Beijing. It can be nearly as polluted at the Great Wall, 70 miles away, as in the city.
32	Will this episode of bad air repeat itself during the Olympics? History says that it probably will, but we shall have to wait and see, because every August is different.
33	As of 1 August, the winds over China are predicted to shift strongly to the south a few days before the Olympics begin.
34	If this prediction holds, it could create bad air for the beginning of the Olympics, because high air pollution in Beijing inevitably comes with air from the south. Nobody wants this for the Olympics.
35	OlympicWatch will be following these predictions very carefully.
36	Comments and predictions Most recent first.
37	Oct 7^th (Tuesday): The API climbed a bit today (to 58 from yesterday’s 25), as anticipated yesterday from the winds returning to the SW. This trend should continue for another day, after which the winds will return to the NW and the APIs will drop.
38	Oct 6^th (Monday): The API dropped further today (to 25 from yesterday’s 50), under the influence of those “temporary” N–NE winds that stayed around a little longer. Now they have returned to the SW, and the API rise tomorrow.
39	Oct 5^th (Sunday): The API dropped again today (to 50 from yesterday’s 64), under the influence of temporary N–NE winds. This effect should not last, however, since S–SW winds are expected to return tomorrow. They should raise the API once more.
40	Oct 4^th (Saturday): The API dropped considerably today (64, vs. yesterday’s 108). This was created by the early arrival of a low-pressure area and rain. It should drop more tomorrow, as the cold front arrives.
41	Oct 3^rd (Friday): The episodic API remained high today (108, vs. yesterday’s 126). It looks now as though it will remain fairly high tomorrow (Saturday), and then crash on Sunday as north winds arrive.
42	Oct 2^nd (Thursday): We are into our first big episode of the fall. The API rose again today (to 126, vs. yesterday’s 104). This is the highest value since we started tracking them on July 1^st. It may not last, though, as winds from the north and northeast are projected to arrive tomorrow. We shall see.
43	Oct 1^st (Wednesday): The API stayed high today (104, vs. yesterday’s 106). From the forecast winds, it looks as though the values should decline tomorrow.
44	Sept 30^th (Tuesday): The API rose rapidly today, to 106 from yesterday’s 84. This high value may mark the beginning of air pollution’s fall season, when strong sawtooth-shaped episodes appear and last from 5 to 10 days. The values at the end, just before the systems crash, can be the highest of the year in Beijing, winter included. Values for the next day or so should decrease, but not by much.
45	Sept 29^th (Monday): Over the weekend, the API rose rapidly, as anticipated on Friday. From Friday’s 17 it reached 84 on Monday, under the influence of southerly flow behind Friday’s cold front. The picture may be unsettled for the next few days, however, because of the big typhoon passing over Taiwan and Japan.
46	Sept 26^th (Friday): The next cold front has arrived, and the API dropped down again (to 17, from yesterday’s 31). But this front will zip past, and the APIs should start to rise again very soon.
47	Sept 25^th (Thursday): The API stayed low today (31, up slightly from yesterday’s 26). A very strong cold front passing through the area should keep it low for another day or so.
48	Sept 24^th (Wednesday): The API could not stay at its extremely low value of 12 from yesterday, and increased to a more usual low of 26. From the forecast winds, it looks as though the API will stay low, and maybe even decrease back to near that 12, during the next day or two.
49	Sept 23^nd (Tuesday): The API decreased today with a vengeance—to 12 from yesterday’s 57. The reason, of course, was the strong cold front that finally arrived and brought strong winds and clean air from the north. This condition may be short-lived, however, as the low is traveling rapidly to the east-northeast. The airflow may become unsettled for a day or so, and then come under the influence of another low that is approaching. We are seeing a foretaste of the winter circulation.
50	Sept 22^nd (Monday): So much for “keeping the API high until at least midday on Sunday,” as predicted yesterday. The API promptly declined to 57 (from yesterday’s 88), giving just one high day. The reason seems to be that the winds strengthened and shifted to the west sooner than expected yesterday. Strong flow from the north should already be beginning, and will last for a couple of days and keep the API low. (Continued on next page)
51	What about the end of the ban? We are now two days past it, and nothing clear seems to be happening. The return of the other 50% of the autos to Beijing’s roads seems be having no detectable effect on the API, the same as happened at the beginning of the ban. It is still early to detect changes, however, and we will keep watching.
52	Sept 21^st (Sunday): The API finally increased significantly today (to 88 from yesterday’s 59), after my having predicted it for the last few days. It was caused by organized flow from the south, which should keep the API high until at least midday on Sunday. That will be followed by a strong cold front around midday on Monday, at which time the API should decrease sharply. (Continued on next page)
53	Does the big increase right after the ban expired mean that the ban was effective? Not necessarily, because the increase also coincided with flow from pollution sources to the south of Beijing. We will have to monitor the API for at least a couple more weeks to get a feeling as to how much the ban really decreased the API.
54	Sept 20^th (Saturday): The API dropped to 59 today (from yesterday’s 67). This was a little earlier than expected, and seems to correspond to the winds becoming unsettled rather than from the south. From the forecast winds, it looks as though the API will still increase Saturday to Sunday, and decrease sharply on Monday.
55	Note in passing: Today, September 20^th, is the end of the big ban on auto traffic and other emissions. We will follow the APIs for long enough to see whether they rise detectably.
56	Sept 19^th (Friday): The API dropped a little today (to 67 from yesterday’s71), for reasons that are not clear from the maps. (These things happen.) The overall projection is much the same as yesterday’s, however—a rise through Sunday, and then a sharp decrease as a strong cold front takes over on Monday.
57	Sept 18^th (Thursday): The API climbed a little today (71, vs. yesterday’s 63), as anticipated yesterday. It may continue to rise through Sunday, but a strong cold front will change all this on Monday.
58	Sept 17^th (Wednesday): The API remained about the same today (63, vs. yesterday’s 64). It will probably stay the same or climb slightly during the next two or three days.
59	Sept 16^th (Tuesday): The API crept up a little today (64, vs. yesterday’s 59), but it didn’t make anything to worry about. It will probably climb steadily for the next two or three days, as winds from the south build in.
60	Sept 15^th (Monday): Again the API remains stable (59, vs. yesterday’s 56). It should stay this way for another day, and then begin to climb on the 17^th (Wednesday), as coherent flow from the south develops.
61	Sept 14^th (Sunday): Again the API remained about the same (56, vs. yesterday’s 58). It should stay this way until the basic meteorological system of lows north and south of Beijing changes. This is not expected for two or three more days.
62	Sept 13^th (Saturday): The API stayed the same as yesterday (58 vs. yesterday’s 59), as anticipated from yesterday’s meteorological predictions. For the next couple of days, Beijing will be caught between the storm over Taiwan and a storm to Beijing’s north. The result will be unsettled circulation and probably unsettled API’s, with no particular trend.
63	Sept 12^th (Friday): The API is just jumping around these days, now up to 59 from yesterday’s 47. It seems that a strong storm over Taiwan is preventing coherent southerly flow from developing and increasing the API at Beijing. This situation should continue for as long as the storm stays there, which will be a couple more days at least.
64	Sept 11^th (Thursday): The API remained about the same (47, vs. yesterday’s 52), because the winds remained generally westerly. The situation should be similar tomorrow, and perhaps move toward southerly flow and higher APIs on Saturday.
65	Sept 10^th (Wednesday): The API decreased a little (to 52 from yesterday’s 65), under the continuing westerly winds that had been anticipated yesterday. Tomorrow may be a little cleaner, as the winds shift more to the northwest. Except for one day at the beginning, the Paralympics have had middle-of-the-road air quality so far.
66	Sept 9^th (Tuesday): The API increased moderately (to 65 from yesterday’s 50) as the winds turned to the W and SW a little sooner than forecast. They should remain generally from the west for another day or two. That means little basic change in air quality.
67	Sept 8^th (Monday): Remember that southerly flow that I said was going to keep the API increasing for another day or so? Well, it got cut off by a low-pressure system descending from the north faster than expected. As a result, the API declined to a respectable 50 from yesterday’s 86. The W-NW winds should continue, with various degrees of disorganization, for another day or two. That means continuing good air for a while.
68	Sept 7^th (Sunday): The API jumped up today (to 86 from yesterday’s 59), as raised yesterday as a possibility because organized flow from the south was anticipated to develop. Similar flow should continue for the next day or two, and should continue to increase the IAPs, at least for the first day.
69	Sept 6^th (Saturday): The API again remained steady (59, vs. yesterday’s 58), as predicted yesterday. In the next day or two, however, it may begin to rise, as organized flow from the south develops.
70	Sept 5^th (Friday): The API remained steady today (58, vs. yesterday’s 57). This reflects irregularity in the winds, with no particular organized circulation dominating. This condition should last for at least the next day or two.
71	Sept 4^th (Thursday): The expected upward trend in API (from Wednesday’s 73) did not continue. Instead, it decreased to 57. The reason seems to be that the wind turned toward the west (and away from the strong sources to the south) sooner than anticipated. These things happen. J
72	The shift was caused by a low-pressure area to the northwest moving toward Beijing a little faster than anticipated. This low pressure area should continue to dominate the airflow near Beijing for the next day or so. That means that APIs should remain steady or decrease.
73	Sept 3^rd (Wednesday): As expected, the API rose significantly today (to 73 from yesterday’s 37), under the influence of organized airflow from the south. This trend should continue through tomorrow, Thursday.
74	Sept 2^nd (Tuesday): Yesterday’s low API of 25 proved to be the end of a two-day minimum, with today’s value rising to 37. The latest wind forecasts suggest that a strong two-day episode is on its way, with very high APIs anticipated for Wednesday.
75	Sept 1^st (Monday): The API remained at 25 today, as predicted, because of winds from the northwest. The winds should become more variable on Tuesday, and allow the APIs to begin to increase once more.
76	Aug 31^st (Sunday): The API plunged to 25 today, under the influence of winds from the north ( = northwest to northeast). They should shift to the northwest on Monday (keeping the APIs low), and then become more variable on Tuesday. That will set the stage for a return of higher pollution.
77	Aug 30^th (Saturday): The high API of yesterday (111) lasted for only a day. Today it is back down to 64, which is roughly normal for this time of year. Yesterday’s high was caused by strong flow from far to the south, in advance of a cold front. Today’s decrease signals the approaching air from the northwest. During the coming days, the values may drop even lower, as northeasterly flow from a coastal storm takes over.
78	Aug 29^th (Friday): The air pollution is back with a vengeance. Today’s API was 111, up from yesterday’s 79, and well into the unhealthful range. This nearly completes the long increase that began on August 15^th, two weeks ago. We are seeing a typical summer cycle in Beijing, albeit one that is lasting longer than normal.
79	Aug 28^th (Thursday): The API is on the march again. Today it increased from 64 to 79. More importantly, the plots show that it has been steadily increasing since its minimum near 30 on August 15^th. This means that Beijing caught a huge break with the weather—the Olympics came just with the best meteorology of the summer, when winds were from the north and mixing heights (dilution of emissions) were the highest.
80	This sharp minimum during the Olympics is becoming more apparent with each day’s data.
81	Remember that emissions and controls have remained constant over the whole Olympic period, and almost certainly for weeks before and after. Thus, all the variations shown in the plots are from meteorology (the influence of changing weather patterns).
82	Aug 27^th (Wednesday): The API dropped slightly (64 to 56), presumably because of northwesterly airflow just to the northwest of Beijing making itself felt.
83	Aug 26^th (Tuesday): The API stayed about the same (64, vs. yesterday’s 68). Airflow was light from the south, but from the northwest just to the north and west of Beijing.
84	Aug 25^th (Monday): Today, the first day after the Olympics, the API climbed sharply (45 to 68). It seems that the southerly airflow predicted for yesterday finally made itself felt. It may not continue much longer, however, as the organized circulation is due to break up soon.
85	Aug 24^th (Sunday): The API again climbed only slightly (42 to 45), and remained in the “very good” range for the men’s marathon and the closing ceremony. The southerly flow of air did not bring as much pollution as it might have.
86	Aug 23^rd (Saturday): The API climbed slightly today (37 to 42), but remained in the “very good” range. Tomorrow, however, the day of the men’s marathon, could be considerably worse, because a deep southerly flow is developing. With some bad luck, pollution could approach unhealthy levels.
87	Aug 22^nd (Friday): The API fell again today (60 to 37), presumably because of rain and the associated cleaner northern air. It is on track to remain low for the rest of the Olympics.
88	Aug 21^st (Thursday): The API has again risen slightly (54 to 60), but is still in the “good” range. As predicted yesterday, it should remain in this general area for the rest of the Olympics.
89	Aug 20^th (Wednesday): The API has risen slightly (42 to 54), but is still in the “good” range. It should remain in this general area for the rest of the Olympics, as no strong flow from the south is expected.
90	Aug 19^th (Tuesday): The API rose a bit today, from 26 to 42, but is still low. The predictions for winds suggest that, with the possible exception of Wednesday, air quality in Beijing will remain good. It now appears that Beijing will get through the Olympics without violating its contract with the IOC.
91	As noted above, this is largely because of help from Mother Nature. It appears that the Chinese were praying and the Mongolian Weather Gods were listening. Congratulations to both!
92	Aug 18^th (Monday): The API stayed low over the weekend (30s and 40s), with blue skies on Friday and drizzle on Saturday. This last week of the Olympics also should be good, with the possible exception of Wednesday the 20^th.
93	All in all, the Olympics have gotten lucky this time. Maybe the Mongolian Weather Gods were listening.
94	Aug 13^th: Right on schedule, the API is increasing, after two low days. Today it is up to 60. It should remain around that value for a couple more days, and then maybe increase over the weekend.
95	Aug 12^th: Another great day for air quality in Beijing. The API is 32. The forecast remains the same—good air for the next few days. Look for things to start to change next weekend (18^th or 19^th).
96	Aug 11^th: With rain and winds from the north to east, Beijing’s API has dropped into the upper 30s. These relatively clean conditions should continue for the next several days.
97	That means that the Olympics will have a good first week. The people who are worried about China possibly breaking its contract with the IOC can relax for a few days.
98	Aug 10^th: The API increased slightly from yesterday (82 vs. 79). I expect it to remain in this general area (70–90) for the next few days, because the wind forecasts do not show any organized flow from the south setting up until late on the 17^th (one week from today).
99	Aug 9^th: As we had predicted, the pollution began to decrease on the day of the opening ceremony (Aug 8^th). As of the 9^th, the daily API had fallen to 79 (from 94 the previous day). This is good news for all concerned. Concentrations should remain moderate for the next week, based on the predicted patterns of airflow.
100	Aug 8^th: The API remained very high (94), essentially identical to yesterday’s 96. This will make the opening ceremony very hazy. The air will probably remain this way until late Sunday.
101	Aug 7^th, evening (morning of the 8^th in Beijing): All the ingredients seem to be in place for continued heavy haze through the opening ceremony. I hope I am wrong, however. The winds are forecast to continue from the south, and the haze is very strong already. This situation may continue into Saturday or Sunday.
102	Aug 7^th: The wind predictions continue to suggest that heavy haze will continue through much or all of the 8^th. It will then moderate a bit for the next days.
103	Aug 7^th: The SE winds are continuing, and the APIs have built into the upper 90s. This is exactly what is expected from weather patterns like this. It remains to be seen whether it will continue into the opening ceremony tomorrow.
104	August 4^th: The pollution seems to be returning with a vengeance. Today’s API of 83 (up from yesterday’s 35) is higher than I had expected. If this rate continues for even one more day, the pollution will enter the unhealthful zone.
105	Note that all the values are clustering together again, as is typical for a strong regional episode.
106	Aug 4^th: The forecast winds are now pointing to a maximum of air pollution on the 8^th, the day of the opening ceremonies. The pattern is supposed to break after that, with better air returning.
107	Contrary to statements that have been appearing in the media lately, there is nothing abnormal here. Cycles and concentrations like this are normal features of August in Beijing.
108	A word about “mist” Certain representatives of the IOC (including its president) are now claiming that the heavy haze in Beijing is not smog, but natural fog or mist caused by “evaporation.” This is simply untrue, and is misleading the public worldwide. They are doing China a disservice by implying that China cannot tell the truth about its pollution.
109	"It is true that when..." It is true that when relative humidities approach 100%, aerosol particles take up water, grow in size, and make the air look hazier. At 100%, the aerosol becomes mist or fog. But under the current weather conditions in Beijing, that will happen only in the early morning around sunrise, and only on some days. After that, the rising temperature lowers the relative humidity, and the particles return to normal. The vast majority of all “smog” seen in Beijing is aerosol particles, not fog or mist.
110	"In the afternoon," In the afternoon, for example, when relative humidities are around 60%, all the haze is aerosol rather than fog or mist. No one is helped by denying this obvious situation. The current air pollution is Beijing is much worse than anyone would like to see, but we must tell the truth about it. I think the Chinese people are willing to do that, and Jacques Rogge and colleagues should allow them to.
111	We all need to wish the Chinese government the best of luck in dealing with this potentially difficult situation.
112	The End