What is the state of the art?

The European Union has through its Framework research programmes over the last decade strengthened the European research in atmospheric chemistry. Measurement techniques and modelling tools have undergone substantial improvements during this time period. Because of the large number of instruments and techniques to be employed this chapter is longer than the recommended two pages.

The DNPH technique

The use of adsorptive sampling of airborne carbonyl compounds on cartridges coated with dinitrophenylhydrazone (DNPH) followed by analysis of the formed hydrazones with high performance liquid chromatography (HPLC) and ultraviolet (UV) detection is a well tested and proven method. With adequate precaution against interference from oxidants in the air (such as e.g. ozone) large volumes (60-600 L) can be sampled and ample sensitivity can be obtained for ambient air measurements. The identification of peaks in the chromatograms depends on the availability of analytical reference material. Thus, it is possible to overlook high molecular weight carbonyls with a weak UV absorption nevertheless present in significant concentrations. For the investigation in the FORMAT project of possible interference from other carbonyls to the formaldehyde measurements by spectroscopic methods we have planned to conduct the DNPH method with HPLC interfaced to ion trap mass spectrometer (HPLC-MS) through an atmospheric pressure ionisation source. This technique is routinely used in our laboratories for the determination of first generation oxidation products of isoprene (methacrolein and methylvinylketone) as important formaldehyde precursors in the Po Valley.

The Hantzsch technique

The demand for sensitive HCHO measurements on mobile platforms and for studies of air chemistry under the influence of advection processes with short time fluctuations led to the development of fast in-situ techniques using liquid phase fluorimetric detection. Investigations in clean air masses, e.g. the remote maritime boundary layer or the free troposphere where HCHO mixing ratios seldom are above 0.5 ppb (Heikes et al, 1996, Mari, et al, 2000) require detection limits well below 0.1 ppb for quantitative analysis (see Cárdenas et al, 2000).

Based on the first developments of fast on line instruments (Lazrus et al, 1988, Kelly and Fortune, 1994) further upgrade of these techniques was done within the last years improving the stability and the time response of the instruments. While the enzyme fluorimetry (Lazrus et al, 1988) offers a highly specific reaction but costly and sensitive enzymes special emphasis was put on the Hantzsch technique as these instruments offer outstanding detection limits with relative simple and cost effective chemistry. This development now lead to commercially available instrumentation for monitoring networks with detection limits below 50 ppt within 90 seconds and research instruments with even better performance (detection limits below 20 ppt) that can be used to quantify also the low mixing ratios expected in the free troposphere.

Spectroscopic techniques

Satellite sensors may observe the free troposphere and the upper part of the boundary layer. In contrast, ground based remote sensing instruments like SOF and MAX-DOAS observe the whole boundary layer where the highest mixing ratios are expected. In addition they are also sensitive to the free troposphere and stratosphere.

Within the FORMAT project two new remote sensing systems will be deployed. The first of which, known as the mobile Solar Occultation Flux (SOF) method, will be deployed which will enable the measurement of HCHO and other species within the planetary boundary layer. This system is comprised of two spectrometers, UV-VIS and FTIR, an active solar tracker and a GPS receiver mounted on a mobile platform. The Solar tracker locks onto the Sun and thus simultaneous UV and IR Solar spectra are recorded, continuously and independent of the motion of the platform. The system will also be able to simultaneously measure additional molecules as CO and NO2. The existing SOF system has so far been applied in measurements of HCl, SO2, HF and ethylene. Although HCHO measurements has so far not been measured with the system a detection limit of 1 ppb over 2 km slant column is anticipated.

The second novel system, known as Multi-Axis-DOAS (MAX-DOAS) is based on a combination of the well known zenith scattered light DOAS and Off-axis DOAS. Using a 2-dimensional CCD chip as a detector it is now possible to simultaneously observe scattered sky light from different viewing directions. The MAX-DOAS-principle has successfully been applied to the observations of enhanced BrO concentrations in the boundary layer in polar regions. Since HCHO absorptions are analysed in a similar wavelength range in the UV as those of BrO it is possible to apply the successfully tested method directly to the observation of boundary layer HCHO. While the MAX-DOAS is not a proven technique for HCHO measurements, it is a promising new method with an estimated detection limit of 1 ppb tropospheric HCHO and a potential to yield limited profile information from the ground.

Satellite measurements

With the GOME instrument on the European satellite ERS-2, formaldehyde measurements from satellite have for the first time become possible. Several independent studies have shown the feasibility of HCHO retrieval from GOME data, and regions with elevated HCHO concentrations have been identified in the data and linked to biogenic isoprene emissions, biomass burning and also pollution. The current HCHO retrieval algorithm has partially been developed in the frame of the European GODIVA project, but a number of open questions remain to be solved to come to more quantitative results. Important issues include improvements of the spectral fitting routines, determination of the atmospheric light path and correction of cloud effects. For satellite measurements, validation of the results is essential, but up to now no attempt has been made to validate HCHO columns retrieved from GOME spectra.

Modelling

Significant improvements in global CTMs have been made during the last 5 years. Chemical schemes with extensive nitrogen and carbon chemistry are included, and recently with on-line calculations of solar photodissociation rates. Several model intercomparisons have been performed, most recently in connection with the 2001 IPCC Third Assessment Report (TAR). The model performances have also been tested against selected surface based and airborne (balloons, aircraft) observations. These comparisons show that although there still are differences between distributions obtained by the models, and for the models to represent some of the observed distributions, the models are able to reproduce the large scale and temporal behaviour of key chemical compounds like ozone and CO.

Which are the outstanding questions?

In spite of the large European and international effort in tropospheric chemistry research during the last decade there are a number of unanswered questions that limit our understanding of the processes that control the concentration and formation of ozone and other oxidants of the troposphere.

Our limited ability to quantitatively explain ozone formation undercuts our ability to predict future ozone changes as well as the prediction of smog episodes. Some of the questions that need to be addressed are listed in the following sections.

Ground measurements

How precise is our current instrumentation for formaldehyde? The different techniques all have their advantages and disadvantages.

• DNPH cartridge methods use different kinds of ozone scrubbers that may affect the measurements.
  
• Wet chemical analysis using fluorimetric techniques although it is very sensitive is suspected to be affected by interferences. While the well known ozone interference can be quantified and subtracted, and no indication for single compound artifacts has been found, the complex mixture of chemical compounds present in photochemical smog can change the individual laboratory-based results.
  
• Spectroscopic techniques like TDLAS may suffer from unresolved absorption features and, like all the instruments that use gas sample cells, it can be affected by line effects in the inlet systems.
  
• Open path spectroscopic instruments like Differential Optical Absorption Spectroscopy (DOAS) do not suffer from inlet loss problems but, as a multicomponent spectrum has to be deconvoluted, the results may depend on the deconvolution process, especially when mixing ratios close to the detection limits have to be estimated. DOAS systems like MAX-DOAS or AMAX-DOAS depend on the accuracy of the radiation transfer model used for data retrieval.
  

Satellite measurements

Satellites offer global coverage, but it is difficult to measure species in the troposphere, especially because of clouds that reflect light back to space, often from a level above the polluted layer. Among the outstanding questions are:

• Accuracy of the airmass factors for satellite measurements
  
• Proper validation of satellite measurements
  
• How do model results compare with satellite measurements? Do current models reproduce the observations on a global scale?
  

Modelling

Analysed meteorological data (winds, temperature, precipitation, cloud distribution and liquid water content) are now available as input data for global CTMs at relatively high resolution T 63 (1.9 x 1.9 degree) or higher. The increase in computer capacity has made it possible to run global CTMs with such high resolution and full chemistry for periods of months to a year. This allows for detailed studies of the atmospheric oxidation process and the interaction of chemical processes and transport processes on different spatial and temporal scales by selecting during specific time periods when extensive comparisons with key chemical compounds in the oxidation process (ozone, NO2, CO, CH2O) can be made. This will help answer important questions like: What chemical compounds control oxidant formation, how efficient is the oxidation process under different atmospheric conditions, how much does long range transport contribute to local pollution, what is the role of emission from natural sources for the oxidant formation?