Heard D.E. (editor) Analytical Techniques for Atmospheric Measurement

Подождите немного. Документ загружается.

52 Analytical Techniques for Atmospheric Measurement

06 08 10 12 14 16 18

0402 06 08 10 12 14 16 18

17 Aug

16 Aug

9 Aug

Time [UT]

OH [10

–3

]OH [10

–3

]OH [10

–3

]

Figure 1.9 OH intercomparison during the 1994 POPCORN ﬁeld campaign. Diurnal concentration

proﬁles of OH measured by FAGE (open circles; scaled by a factor of 1.09) and DOAS (solid squares)

on 9, 16 and 17 August 1994. Local noon is 11:17 UT. (Taken with permission from Hofzumahaus

et al., 1998.)

OH data gave a constrained fit of slope 1.5 (within combined uncertainties). A calibration

problem associated with these measurements was suggested – not surprisingly, given the

difficulty of calibrating these instruments in flight.

1.7 Atmospheric chemistry and policy

1.7.1 Health effects and environmental policy

For some trace species, emissions and concentrations are subject to a number of national

and international agreements, and are primarily based on health effects. In the UK three

species are most closely targeted and regulated: NO

and particulate matter. At high

Field Measurements of Atmospheric Composition 53

concentrations, NO

causes inflammation of the respiratory airways, and there is evidence

that long-term exposure may affect lung function and that for some individuals exposure

enhances the response to allergens (Pilling, 2004). The European Union (EU) Directive

has an hourly averaged objective of 200 gm

−3

, with 18 exceedances per year, and an

annual average objective of 40gm

−3

.NO

also contributes to the formation of ozone and

of secondary particles and is a contributing factor for acidification and eutrophication.

Under the EU Directive, the UK is committed to reducing its national annual emission of

from 1728 kTonnes as NO

in 2000 to below 1167 kTonnes (Pilling, 2004) by 2010.

Ozone is harmful to vegetation, with a critical level of 40 ppb over a growing season,

and has a damaging effect on certain materials, for example polymers, rubber and

textiles, and is the most common irritant of the common air pollutants and exposure to

high concentrations produces inflammation of the respiratory tract, and at even higher

concentrations is associated with chronic decline in lung function. For O

, the standard

is 50 ppb for an 8-hour running average.

For particulate matter, the diameter is the key parameter controlling the penetration

into the lungs. The metric used in the UK is PM

, which describes the mass of particles

with a size less than 10 m. Health effects resulting from both short-term and long-term

exposure are linked mainly to respiratory and cardiovascular disorders. The EU limit for

is a concentration of 50gm

−3

for a 24-hour day with up to 7 exceedances per

year by 1 January, 2010, with the annual limit being 40 gm

−3

(by 1 January, 2005)

decreasing to an indicative limit value of 20 gm

−3

by 1 January, 2010 (Pilling, 2005).

1.7.2 Monitoring networks

In order to monitor trends and exceedances in air quality standards, and also in long-term

trends in ozone-depleting and greenhouse gases, a variety of ground-based measurement

networks have been set up worldwide on a range of spatial scales. Good accuracy and

especially good precision are important. With the advent of measurements of trace

species in the troposphere from satellites (Section 1.4.8), global coverage is now becoming

possible, albeit with large grid-sizes at the surface (although this is improving dramat-

ically), and with limited temporal coverage during a diurnal cycle because of a single

overpass time and the presence of clouds.

1.7.2.1 Air quality networks

Automatic networks are operated at both the national and the regional scale. In the USA,

the Environmental Protection Agency (EPA) http://www.epa.gov/ operates monitoring

networks for several trace species, including CO, O

,NO

,SO

, dioxins, and PM

and

25

particulates, and UV radiation, in some cases from hundreds of sites. For PM

25

there are anion, cation and carbon analyses using ion chromatography and other methods

from filter samples. Some data are available on-line (e.g. for O

and particles) together

with a national outlook, and the EPA is legally obliged to issue final designations for

all areas of the country stating that they either meet or do not meet the 8-hour ozone

standards. The EPA funds the Clean Air Status and Trends Network (CASTNET), which

54 Analytical Techniques for Atmospheric Measurement

has operated since 1987, to measure dry acidic deposition and rural, ground-level ozone,

and together with other monitoring networks it is able to evaluate the effectiveness of

emission control strategies. The CASTNET consists of over 80 sites, and includes weekly

measurements of sulphate, nitrate, ammonium, sulphur dioxide, and nitric acid, and

hourly measurements of O

concentrations.

In the UK, funding comes from the Department of the Environment, Food and

Rural Affairs (DEFRA), as well as from local authorities, for monitoring networks to

measure CO, ozone, NO

(and NO

), particulate matter (PM

), SO

and hydrocarbons.

CO preferentially combines with haemoglobin in the blood, reducing its oxygen-carrying

capacity, and leading to headaches, depressed heart and breathing rates and death at high

levels.

A list of monitoring sites in the UK Automatic Urban and Rural Network can be found

on the website http://www.stanger.co.uk/siteinfo/Default.asp, and also on the air quality

archive http://www.airquality.co.uk/archive/index.php, with up-to-date information on

current and historic air quality. The number of sites increased from 74 at the end of

1996 to 120 by March 2004. Rural sites mainly monitor O

but also SO

,NO

and PM

in some cases. The DEFRA also funds a hydrocarbon network (35 sites) monitoring

hourly concentrations of 25 volatile organic compounds, including two carcinogenic

compounds, benzene and 1,3 butadiene, in urban roadsides, urban background and rural

locations across the country (Fowler et al., 1997). Hourly data on ambient SO

,NO

, CO and particulate matter concentrations are used to provide air pollution infor-

mation to the public each day. NO

concentrations are measured using diffusion tubes

and chemiluminescence analysers (Chapter 7). For a study in 2001, the UK Air Quality

Expert Group considered NO

measurements from about 200 automatic monitoring sites

(Pilling, 2004). The number of sites had expanded significantly in order to meet the

requirements of EU directives. Many of these sites also have co-located O

chemilu-

minescence analysers. In addition, there is a large-scale survey of NO

concentrations

using passive diffusion tubes, involving over 1300 sites operated by more than 300 local

authorities.

The Tapered Element Oscillating Microbalance (TEOM, Chapter 6) is the most

commonly used instrument used for PM

monitoring. This continuous sampler provides

real-time data with a short time resolution (<1 hr) and a precision of ±2gm

−3

, but

a disadvantage is that it incorporates a heated manifold, which may lead to losses of

semi-volatile species such as ammonium nitrate. Hence a correction is necessary when

comparing with measurement of PM

using other methods. Inside the instrument is a

tapered glass tube, whose frequency of oscillation depends upon its mass. Any change in

the mass of the tube, due to the deposition of particles onto a small filter affixed to one

end, will result in a change in the resonant frequency, with the change being proportional

to the additional mass. The instrument uses a size-selective inlet (see Chapter 6 for details

of inlets used for aerosol measurements). There are 240 PM

monitoring sites in the UK,

15 of which have co-located PM

25

analysers, and measurements are also made of black

smoke, polycyclic aromatic hydrocarbons (PAHs), heavy metals (lead, nickel, arsenic and

cadmium) and major ions (such as sulphate, nitrate and chloride).

Measurement uncertainty is very important if data are to be used to monitor health

damaging pollutants and to enforce/influence legislation. The objectives set down by EU

directives for NO

are 15% for continuous measurements (e.g. using chemiluminescence

Field Measurements of Atmospheric Composition 55

analysers) and 25% for indicative measurements (e.g. using diffusion tubes), at the 95%

confidence level for individual measurements averaged over the limit period (e.g. 8 hr).

Quality control arrangements ensure that measurement precision standards of at least

±10 gm

−3

for SO

, ±8 gm

−3

for NO

and O

, ±05mgm

−3

for CO and ±4 gm

−3

for PM

are achieved at automatic monitoring sites. Further details on uncertainties of

particular techniques are given in later chapters. Industry is under increasing pressure to

clean up its exhausts, under penalty of heavy fines for non-compliance, and many gas

and particulate analysers are employed to monitor stack emissions.

1.7.2.2 Networks for long-term measurements

of stratospheric ozone depleting substances,

greenhouse gases and tropospheric ozone precursors

Air quality networks tend to be regional and within the same country. There are also

a number of monitoring networks that transcend continental boundaries and which

have been operating for over thirty years. These networks document global change,

provide data to evaluate ozone depletion and climate change, enable the predictions of

models to be tested, and inform policy makers. Finally, with the advent of long-term

measurements of tropospheric trace species from satellites, the ground-based networks

provide an important validation service.

The World Meteorological Organization has set up the Global Atmospheric Watch

(GAW) programme, which coordinates networks to measure ozone (total column,

tropospheric, vertical profiles), greenhouse gases, solar radiation, aerosol character-

istics and tropospheric ozone precursors (CO, NO

). The NOAA Climate Monitoring

and Diagnostics Laboratory (CMDL, http://www.cmdl.noaa.gov/) programme supports

measuring sites all around the world, both on the ground and using tall towers and

aircraft platforms. Sites are located at Point Barrow, Alaska (71



N, 156



W), Trinidad

Head, California (41



N, 124



W), Mauna Loa (195



N, 156



W, 3397 m above sea level),

American Samoa (14



S, 171



W) and the South Pole (90



S, 102



W). As well as in situ

instruments, whole air samples taken using flasks at a range of sites, all away from

local influences, are shipped back to the USA for analysis. Surface measurements have

revealed distinctive seasonal and latitudinal patterns, as well as long-term trends, for

several species, for example CO, CO

,CH

O, O

(tropospheric and stratospheric),

hydrocarbons, halogenated VOCs, aerosols and radiation. Another network consisting

mainly of in situ instruments is the Atmospheric Lifetime Experiment/Global Atmospheric

Gases Experiment ALE/GAGE, which has been measuring CH

O, several CFCs and

other chorine compounds using gas chromatography (Chapter 8) since 1978 [Prinn

et al., 1983]. The advanced GAGE (AGAGE, http://agage.eas.gatech.edu/) now uses fully

automated instruments. The AGAGE monitoring sites are Mace Head (Ireland, 53



W), Cape Grim (Tasmania, 41



S, 145



E), Trinidad Head (California, 41



N, 124



W),

Ragged Point (Barbados, 13



N, 59



W) and Cape Matatula (American Samoa, 14



S, 171



W),

with data also obtained at Ny-Ålesund (Svalbard, 789



N, 119



E) and the Jungfraujoch

(Switzerland, 46



N, 8



E, 3580 m above sea level). Another ground-based network is the

Network for Detection of Stratospheric Change (NDSC), which monitors stratospheric

change (temperature, O

, other trace species and radiation) using a range of remote

sensing instruments. A further example is the GAW ozonesonde and CO

network.

56 Analytical Techniques for Atmospheric Measurement

1970

1975 1980 1990 1995 2000 20051985

Year

320

340

360

380

400

420

(ppm)

1900

1800

1700

1600

1500

1400

(ppb)

Barrow CO

Mauna Loa CO

Samoa CO

South Pole CO

Barrow CH

Mauna Loa CH

Figure 1.10 Concentrations of the greenhouse gases methane and CO

measured by the NOAA CMDL

monitoring network since 1973. Methane is measured using a gas chromatograph with a ﬂame ionisation

detector and CO

is measured using a non-dispersive infrared (NDIR) analyser. (Data courtesy of Pieter

Tans, NOAA.) (Reproduced in colour as Plate 2 after page 264.)

Figure 1.3 shows the long-term trend in concentrations of CFC-11 CFCl

 and CFC-12

(CF

), and Figure 1.10 shows measurements of CO

and CH

from 1973 to 2005,

taken at a variety of remote, background sites. CFCl

(atmospheric lifetime of 45 years)

was banned following the 1987 Montreal Protocol, and as shown concentrations are now

beginning to fall off. However, concentrations of CF

have been slower to respond

since their ban following the Montreal Protocol, due to the longer atmospheric lifetime

(100 years) of this species. However, the loading of chlorine in the stratosphere will not

fall below the critical value for formation of the ozone hole for at least another fifty years

or so, according to predictions from a range of global climate models.

Although not a monitoring network, Berresheim et al. (2005) have made a study of

the long-term behaviour of the OH radical over five years at the Hohenpeissenberg

observatory in Southern Germany, using a chemical ionisation mass spectrometer (see

Chapter 5). In this way a seasonal study of the behaviour of this radical has been possible.

The results indicate a very close correlation between OH radical concentrations and the

rate of OH production (P(OH)) from the rate of ozone photolysis to generate O

D

atoms (referred to as JO

D, see Chapter 9). A linear relationship is expected from

Equation 1.3, with the slope of the correlation changing as the chemical lifetime of OH

(by reaction with its sinks) changes when there is a change in the sampled air mass,

perhaps bringing in more polluted air from urban areas.

Monitoring of long-lived anthropogenically emitted species (e.g. methyl chloroform,

CCl

), whose loss rate is controlled almost totally by OH radicals has been

used to infer globally averaged OH concentrations, although the value obtained

Field Measurements of Atmospheric Composition 57

(11 ×10

molecule cm

−3

) is weighted towards the tropics, where most of the CH

CCl

is oxidised. Measurements of CH

CCl

between 1978 and 1994 did not suggest any

significant trend in [OH], which suggests that production rates of OH during this period

must have increased (e.g. UV levels due to reduced ozone in the stratosphere), as the

concentrations of two key sinks of OH – CO and CH

– increased during this period.

However, concentrations of CH

CCl

are falling year-on-year, as it is a banned ozone

depleting substance following the Montreal Protocol, and it will not be easy to infer any

future long-term trends in OH concentrations, and hence the oxidising capacity of the

atmosphere, from future measurements of CH

CCl

1.8 Major ﬁeld campaigns for measurement

of atmospheric composition

Major intensive field campaigns are normally mounted over a period of 1–3 months. The

duration maximises the probability of experiencing the desired meteorological conditions

(fine weather) to achieve the scientific objectives. For a ground-based experiment, the past

history of the sampled air mass, the range of emissions the air has encountered over the

previous five days or so (viewable using back trajectories), is the most critical parameter,

as it controls the mix of trace gases. Often particular wind directions are undesirable as

they allow local sources to interfere with the experiment. For a Lagrangian study, when

it is desired to study the evolution of composition within a moving air mass, perhaps

between a land-based observatory and a ship or aircraft, or between two or more aircraft

over a very wide geographical area, the opportunity for connected flow is quite limited,

and much use is made of forecast modelling. Process studies are often best studied using

intensive field campaigns involving large number of instruments and scientists from

national or international consortia. The advantage of this approach is that models can be

constrained with a large number of input parameters, and it is possible to study several

research strands at the same site. One missing measurement can endanger an entire

science objective, and instrument redundancy should be built into field campaigns, which

are expensive to operate. Even if only two instruments for measurement of the same

species work perfectly, confidence in that measurement is higher as an intercomparison

can be performed.

1.8.1 The design of ﬁeld campaigns

It is now accepted that the concerted deployment of a suite of field instrumentation is

necessary to provide a detailed understanding of the chemical composition for a given

environment. Naturally, it is not possible to measure every conceivable trace species,

and campaigns must be designed to cover the necessary measurements to deliver a

particular science objective. Modelling plays a key role over a range of length scales, from

processes at local scale to long-range transport. There are examples of field sites where

multiple intensive campaigns have been mounted over a period of several years, with

each successive campaign becoming more and more comprehensive in terms of species

coverage, as technology evolves and new laboratory techniques make it into the field.

58 Analytical Techniques for Atmospheric Measurement

During the Eastern Atlantic Spring Experiment at Mace Head in 1997, OH and HO

measurements were made using LIF (see Chapter 4), and compared with the predictions

of zero-dimensional model. Despite the model being constrained with a large number of

sources and sinks of HO

, and the most up-to-date kinetics and photochemical data being

used, the model consistently overpredicted both species, sometimes by up to a factor of

1.5 to 2, suggesting that important sinks of each species were not being measured. Upon

return to Mace Head in 2002 for the North Atlantic Marine Boundary Layer Experiment

(NAMBLEX), a wider range of species was measured, notably oxygenated volatile organic

compounds (OVOCs) – methanol, acetone and acetyaldehyde – and halogen oxides

(IO and BrO). The concentrations of OVOCS were found to be removing a larger fraction

of OH molecules than non-methane hydrocarbons, and indeed the agreement of the

model with measured OH was much better once OVOCs were included. For HO

, the

levels of measured IO and BrO were found to remove a significant fraction of HO

, and

model–measurement agreement was better when these reactions were included. With the

benefit of hindsight, it became clear that measurements of key species necessary to fully

describe the chemistry were missing in the previous campaign.

The requirements of a measurement platform and instrument portfolio depend on

the precise aims of the study. If CO

or methane concentrations representative of the

background atmosphere are required to assess any trends in their contribution to global

warming, then a ground-based field site remote from any major emission sources of these

compounds is needed. Ideally, in order to provide consistency checks and to rule out

local sources, a network of such stations is needed (see Section 1.7.2). If the objective

is to understand the destruction of ozone in the lower stratosphere over a range of

geographical locations, then probably one would choose a combination of remote sensing

methods with good vertical resolution (e.g. airborne LIDARs or satellite instruments)

with in situ instruments aboard high flying aircraft, (e.g. the ER-2) to measure ozone

and a range of short-lived free-radical species and their longer-lived precursors, together

with aerosol species (e.g. polar stratospheric clouds).

The number of field campaigns mounted is now very large indeed, and it is out of the

scope of this book to try to attempt a comprehensive compilation or discuss them in any

detail. In a 2003 review by Heard & Pilling, 46 major field intensive campaigns, between

1991 and 2001, which involved tropospheric measurements of OH and HO

radicals

were listed. Table 1.3 provides a very selective list of some of the major field campaigns

that have taken place in the last five years, and is included to illustrate the wide variety

of platforms and geographical locations where measurements have been taken. There

are a number of special issues in journals dedicated to the results from intensive field

campaigns, and some of these are listed in ‘Further reading’. For illustrative purposes,

the details of one field campaign are given in Section 1.8.2.

The volumes of data created during a comprehensive field campaign are truly enormous,

and if the project is to succeed, there must be an agreed mechanism, and deadline, for

depositing field data. All large field campaigns should have a central database. In the UK,

all fieldwork data for atmospheric composition are managed by the British Atmospheric

Data Centre (BADC). The BADC is centrally involved at all stages of the field project

from its inception, and facilitates the agreement of data protocols to protect intellectual

property rights, with data normally released to the general public after some agreed delay

following the end of the project. The Centre also provides the tools, technical support

Field Measurements of Atmospheric Composition 59

and secure environment for on-line deposition of data, using an agreed, common format,

usually NASA Ames or Net-CDF, and is responsible for setting up password-protected

workspace areas where data can quickly be deposited by project participants and shared

with other users prior to the formal delivery of validated data to the central archived

database.

1.8.2 Case study of a ﬁeld campaign: The 2002 NAMBLEX

campaign

In the summer of 2002, ca. 50 scientists from 12 institutions participated in the NAMBLEX

campaign, at Mace Head 53



N 10



W, on the West Coast of Ireland. Mace Head

experiences mainly unpolluted air from the Atlantic Ocean brought in by the prevailing

westerly winds. The site is part of AGAGE and is a GAW Station (Simmonds et al.,

1996). The aims of NAMBLEX were to test quantitatively the basic understanding of

oxidation processes in clean and moderately polluted air using comparisons of measured

and model-predicted free-radicals, to study extensively the chemistry of halogen species

in the marine boundary layer (MBL) through observation of reactive intermediates and

their sources and sinks, to examine the origins and role of reactive hydrocarbons in the

MBL, and to investigate the size-distributed composition of aerosols and the processes

involved in new particle creation. To achieve these aims a wide range of gas phase and

aerosol species was measured using a range of state-of-the-art instrumentation, as shown

in Table 1.4, which also lists averaging times and detection limits. Wherever possible,

inlets were positioned at the same height above the ground (10 m is often used), and

their horizontal displacement was minimised. Figure 1.6 shows a photograph of the

main field site at Mace Head during NAMBLEX. The majority of the instruments were

deployed in bespoke converted shipping containers, of standard size for convenience of

transportation by road or sea. These mobile laboratories are self-sufficient, requiring only

a source of electricity. Some instruments were also housed in buildings adjacent to the

containers. Measurements of several free-radical species were made, namely OH, HO

,NO

, BrO and IO (see also Table 1.1 for methods of their detection), together with

their sources and sinks, and instrument inlets were placed as close as possible. For some

species (e.g. CH

O, peroxides, NMHC) measurements were made using more than one

technique, enabling an intercomparison to be performed (see Section 1.6.3).

Chemists, physicists and meteorologists worked together closely during NAMBLEX.

Detailed chemical compositional measurements ran alongside physical measurements of

the structure of the boundary layer, using, for example, a small tethered balloon equipped

with a basic met package (with telemetry), a SODAR (sonic detection and ranging)

and a wind-profiler (1290 MHz Doppler radar) for vertical distributions of wind speed

and direction, and several sonic anemometers mounted at various heights on masts, to

measure the u, v and w wind components in real time. Combining the measurements

from the wind profiler, SODAR and sonic anemometers allow the validity of trajectory

calculations to be tested. There were several events where the local wind direction at the

site was decoupled from the synoptic long-range flow, as represented by five-day air mass

back-trajectories, which indicate the trajectory of the air sampled at the site during the

previous five days. Air mass trajectories were calculated using wind fields modelled by

60 Analytical Techniques for Atmospheric Measurement

Table 1.4 Species measured and instruments deployed during the North Atlantic Marine Boundary Layer

Experiment (NAMBLEX), Mace Head, Ireland, held during the summer of 2002

Species

measured

Analytical technique Temporal

resolution

Detection limit Chapter

OH Fluorescence Assay by Gas

Expansion (FAGE)

30 s 2 ×10

molecule

−3

(150 s),

0.002 pptv

Addition of NO to convert to OH,

then FAGE

30 s 1 ×10

molecule

−3

(150 s)

0.04 pptv

4, 7

RO

+HO

Peroxy Radical Chemical Ampliﬁer 1 min 0.4 pptv (10 min) 7

, IO, OIO,

BrO, I

Differential Optical Absorption

Spectroscopy, optical path

2 ×42km

5 min 0.5, 0.8, 5,

2, 5 pptv

, OIO, I

Broad-band cavity ring-down

spectroscopy

100 s–10 min 1, 4, 20 pptv 3

Vertical proﬁles

of NO

Zenith-pointing DOAS and altitude

retrieval

1 proﬁle/day 1–5×10

molecule

−2

Photolysis

frequencies,

including

JO

D,

JNO



Calibrated ﬁlter (2 sr and 4 sr)

radiometers and scanning

spectral-radiometer

1s 9

UV absorption spectroscopy 1 s 0.6 ppbv 3

NO NO/O

chemiluminescence

detectors

10 s 1 pptv 7

Photochemical convertor

→ NO

2 pptv 7

NO Gold tube/CO converter

→ NO

5 pptv 7

NO

-HNO

Gold tube convertor/denuder to

remove HNO

10 pptv 7

,CH

OOH Scrubbing/derivatisation, high

performance liquid

chromatography

8 min 10 pptv 7

,  ROOH Dual-channel ﬂuorometric detector 20 s 10 pptv 7

O vapour IR absorption 10 s 5 ppmv 2

Broad band cavity ring-down

spectroscopy

O Fluorometric detection (Hantzsh

reaction)

1 min 20 pptv 7

Selective trapping and gas

chromatography / helium

ionization detection

15 min 50 pptv 8

Field Measurements of Atmospheric Composition 61

Table 1.4 (Continued)

, speciated

-C

non-methane

hydrocarbons,

including

isoprene, DMS

Automated GC/ﬂame

ionisation detection

using programmed

temperature injection

Seawater sampling using

head-space gas

chromatography

40 min 1 pptv 8

OH,

CH



CO,

CHO

Gas chromatography with

ﬂame ionisation detection,

selective column for

oxygenates

40 min 1 pptv 8

Peroxy acetyl

nitrate (PAN)

Gas chromatography/electron

capture detection

5 min 10 pptv 5

Alkyl nitrates Negative ion chemical

ionisation gas

chromatography/mass

spectrometry

45 min <1 pptv 8

Reactive

halocarbons

Gas chromatography/mass

spectrometry

40 min <1 pptv 5

Aerosol number

and size

distributions

(3 nm to 20 m)

Scanning mobility particle

sizer (3–500 nm), range of

optical particle counters

01–300 m, CN counters,

>3, 5, 10 nm)

1 min 6

Hygroscopic

properties

(growth factors)

for 20–300 nm

aerosol particles

Hygroscopic tandem

differential mobility analyser

5 min 6

Aerosol chemical

composition

Aerosol mass spectrometer

(Aerodyne Inc.)

0.05 s 100 particles s

−1

(025 gm

−3

for

a 3 min average

Aerosol time of ﬂight mass

spectrometer (TSI Inc.)

Cascade impactors (ﬁlters)

with ion chromatography

(NH

,Na

,Cl

−

,SO

2−

)

12 hr 6,8

Vertically resolved

wind-speed and

direction u, v,

w wind velocity

Wind proﬁler (Doppler

radar), Sodar, tethered

balloon

15 min 1

Sonic anemometers 0.1 s 1

CO Gas chromatography/hot

mercuric oxide reduction gas

detector coupled with UV

detection

30 min 8