Fig' 3 Ozone mole fraction from ozone monitor in red, relative humidity RH in blue, wind direction i

1
AGU Abstract No. A11F-0063 A Preliminary Field
Campaign of the Kathmandu Valley, Nepal An Urban
Photochemistry Study Elke Hodson1, Arnico
Panday1, Yong Yu2, Ronald Prinn1, Bo
Galle2 1Department. of Earth, Atmospheric, and
Planetary Sciences, Massachusetts Institute of
Technology, 54-1413, Cambridge, MA 02139, USA
2Department of Radio Space Science, Chalmers
University of Technology , 412 96 Gothenburg,
Sweden Email elkeh_at_mit.edu

• WHY THE KATHMANDU VALLEY?
• Located in a semi-enclosed basin at high altitude
  severe winter temperature inversions
• Rapidly growing population and vehicle ownership
  per capita
• Less is known about the mechanism of ozone
  formation in developing countries (Brasseur et
  al., 1998).
• To our knowledge, ozone and NO have never been
  measured in Kathmandu.
• In order to better characterize the fast
  photochemistry of Kathmandu and to test several
  field instruments, a pilot field campaign was
  conducted in January-February 2003.
• Figure 1. Northeast Kathmandu, Nepal.
  Research site is at wingtip. Downtown is 2
  km west from research site. A major road is
  100m south of research site. Photo taken by
  Arnico Panday in November, 2002

• CONCLUSIONS
• Ozone and NOx pollution are less of a health
  problem than aerosol loading.
• Typical urban pollution trends are observed for
  O3, NO, and NO2.
• Ozone mole fractions are more dependent on
  meteorology than NOx mole fractions.
• More work needs to be done to characterize
  horizontal and vertical variability of these
  trace gases and their dependence on
  meteorological parameters.

What does ozone look like on a typical winter day
in Kathmandu?
Jan 29
Feb 3
Fig. 3 Ozone mole fraction (from ozone monitor)
in red, relative humidity (RH) in blue, wind
direction in black, and solar irradiance in green
for Jan 29 and Feb 3. The top two panels are for
Jan 29 and the bottom three for Feb 3. High RH,
low solar irradiance, and low erratic winds
indicate low SL-ML height. See Fig 5 for wind
speeds. Jan 29 and Feb 3 were the lowest and
highest ozone mole fractions recorded by the
ozone monitor, respectively. For an explanation
of the trends in this figure see text box below.
Fig. 2 Ozone data from Jan. 26-Feb. 6, 2003 from
both the DOAS instrument (black scattered dots)
and ozone monitor (red line). The x axis is in
Julian days with Jan 1, 2003 being day 1. Each
subplot represents three days. In general, ozone
peaks at or just after midday with a sharp
increase occurring just before noon. This is
followed by either 1) a slow decrease until
midnight when the surface boundary layer (called
SL-ML) remained low or 2) an initial sharp drop
followed by a slower decrease when the SL-ML kept
growing even after photolysis peaked.

• What is the reason for the diurnal ozone
  variability?
• High daytime surface ozone mole fractions are
  caused by entrainment of air with high ozone mole
  fractions from above the nocturnal boundary layer
  (SL-ML) and high photochemical production (P)
  within the SL-ML (e.g. model study not shown
  here, Fast et al, 2000 Berkowitz et al, 1997).
  High entrainment is indicated by early and high
• SL-ML growth. The difference in maximum ozone
  mole fractions achieved on Jan 29 vs. Feb 3 is
  due to a combination of these two factors (see
  Fig 3).
• Jan 29
• low SL-ML during day (solar irradiance low, RH
  high, steady afternoon winds for only a few
  hours)
• low P (low solar irradiance)
• Feb 3
• SL-ML grew early and high (solar irradiance high,
  RH low, stable northwesterlies throughout
  afternoon)
• strong P (high solar irradiance)

• EXPERIMENTAL
• The major instruments included
• Ozone monitor (2B Technologies)
• NOx (NO NO2) Analyzer (Model 42S Thermo
  Environmental Instruments Inc.)
• DOAS (Differential Optical Absorption
  Spectroscopy) instrument (Chalmers)--measured O3,
  NO2, HONO, CH2O, SO2
• Weather station (Rainwise) -- measured T, P, RH,
  u, v, solar irradiance

Fig. 5 For each set of 3 days, NOx is shown first
with wind direction shown underneath. NO from
NOx analyzer, NO2 from DOAS. The time scale for
each set of NOx and wind vectors is the same.
Line slope indicates the direction the wind is
heading. NOx profiles look very different than
those for ozone. In general, NO shows a clear
morning peak during the morning rush hour. The
morning NO peak does not seem to correlate
strongly with wind direction and wind strength.
An evening rush hour does occur, but ozone mole
fractions are still so high that most emitted NO
is probably lost by reaction with ozone. Most
likely the observed NO spikes are from nearby
vehicular or biomass burning. Southerly winds
would correspond to polluted air blowing in from
the major road to the south. Downtown is
west/southwest. Several days show a prolonged
period of stable northwesterlies in the
afternoon. This does not seem to correspond with
lower observed morning NO maxima, due to
dilutions from a fast growing SL-ML.
Fig. 4 Noon of Feb 12 - noon of Feb 13. Ozone
from ozone monitor NO and NO2 from NOx analyzer.
Only day where NO2 channel and ozone instrument
were working simultaneously. This day seems to
have been a large pollution event. Both NO and
NO2 show mole fractions larger than any observed
in Fig 5. The high frequency NOx measurements
help resolve the morning NO2 trend. NO (blue)
and NO2 (black) begin to increase in the morning
at about the same time. NO rises faster to
higher mole fractions than NO2. Then as ozone
mole fractions rise due to an influx of ozone
aloft and increased photochemical production in
the BL from NO2 photolysis, NO is depleted first,
followed soon after by a decrease in NO2. NO2
shows a morning and evening peak because unlike
NO, NO2 has no appreciable evening photochemical
loss pathway.

• References
• Berkowitz, C. M. and W. J. Shaw, J. Geophys.
  Res.-Atmos., 102(D11), 12795-12804, 1997.
• Brasseur, G. P., J. T. Kiehl, et al., Geophys.
  Res. Let., 25(20), 3807-3810, 1998.
• Fast, J. D., J. C. Doran, et al., J .Geophys.
  Res., 105(D18), 22833-22848, 2000.